Transform your 3D texturing workflow with the power of generative AI, directly within Blender!
feat: pbr auto-saturation, texture prompt, artifact fail, showcase overhaul
sakalond9ec6ca0a3519
diff --git a/README.md b/README.mdindex 26cb6ab..226c1df 100644--- a/README.md+++ b/README.md@@ -4,9 +4,9 @@[](#system-requirements)[](https://github.com/sakalond/stablegen/releases)-**Create 3D assets from images and prompts, then texture and refine them — all inside Blender.**+**Create 3D assets from images and prompts, then texture and refine them - all inside Blender.**-StableGen is an open-source Blender addon that brings generative AI into your 3D workflow. **Generate** fully textured 3D meshes from a single image or text prompt via [TRELLIS.2](https://github.com/microsoft/TRELLIS.2), then **texture and refine** them — or any existing model — using SDXL, FLUX.1-dev, or Qwen Image Edit through a flexible [ComfyUI](https://github.com/comfyanonymous/ComfyUI) backend.+StableGen is an open-source Blender addon that brings generative AI into your 3D workflow. **Generate** fully textured 3D meshes from a single image or text prompt via [TRELLIS.2](https://github.com/microsoft/TRELLIS.2), then **texture and refine** them - or any existing model - using SDXL, FLUX.1-dev, or Qwen Image Edit through a flexible [ComfyUI](https://github.com/comfyanonymous/ComfyUI) backend.---@@ -19,6 +19,8 @@ StableGen is an open-source Blender addon that brings generative AI into your 3D- [💻 System Requirements](#-system-requirements)- [⚙️ Installation](#️-installation)- [🚀 Quick Start Guide](#-quick-start-guide)+ - [Texturing an Existing Model](#texturing-an-existing-model)+ - [Generating a 3D Model with TRELLIS.2](#generating-a-3d-model-with-trellis2)- [📖 Usage & Parameters Overview](#-usage--parameters-overview)- [📁 Output Directory Structure](#-output-directory-structure)- [🤔 Troubleshooting](#-troubleshooting)@@ -54,12 +56,12 @@ StableGen brings AI-powered 3D generation and texturing directly into Blender:* Sophisticated weighted blending ensures smooth transitions between views.* 📷 **Advanced Camera Placement:*** **7 placement strategies:** Orbit Ring, Fan Arc, Hemisphere, PCA-Axis, Normal-Weighted K-means, Greedy Occlusion Coverage, and Interactive Visibility-Weighted placement.- * **Per-camera optimal aspect ratios** — each camera gets its own resolution computed from the mesh's silhouette, so no pixels are wasted on letterboxing.- * **Unlimited cameras** — no more 8-camera limit.- * **Camera generation order** — drag-and-drop reorder list with 6 preset strategies to control the processing order in Sequential mode.+ * **Per-camera optimal aspect ratios** - each camera gets its own resolution computed from the mesh's silhouette, so no pixels are wasted on letterboxing.+ * **Unlimited cameras** - no more 8-camera limit.+ * **Camera generation order** - drag-and-drop reorder list with 6 preset strategies to control the processing order in Sequential mode.* Camera cloning, mirroring, and floating viewport prompt labels.* 🎯 **Local Edit Mode:**- * Point cameras at specific areas to modify — new texture blends seamlessly over the original using angle-based and vignette-based feathering.+ * Point cameras at specific areas to modify - new texture blends seamlessly over the original using angle-based and vignette-based feathering.* Separate angle ramp and silhouette edge feathering controls for precise blending.* Works with all architectures (SDXL, Flux, Qwen Image Edit).* 📐 **Precise Geometric Control with ControlNet:**@@ -96,9 +98,94 @@ See what StableGen can do!----### Showcase 1: Head Model Stylization with IPAdapter+### Showcase 1: Text-to-3D (SDXL)-This showcase demonstrates how StableGen can texture the model using a standard prompt and then with style guidance from an IPAdapter image reference.+Assets generated entirely from a text prompt using the TRELLIS.2 pipeline with SDXL-based texturing.++| | | | |+| :------: | :------: | :------: | :------: |+| <img src="docs/img/trellis2/sdxl_dragon.gif" alt="Fantasy dragon" width="200"> | <img src="docs/img/trellis2/sdxl_wizard.gif" alt="Wizard character" width="200"> | <img src="docs/img/trellis2/sdxl_hut.gif" alt="Hut" width="200"> | <img src="docs/img/trellis2/sdxl_telescope.gif" alt="Telescope" width="200"> |+| *Dragon* | *Wizard* | *Hut* | *Telescope* |+| <img src="docs/img/trellis2/sdxl_robot.gif" alt="Robot" width="200"> | | | |+| *Robot* | | | |++<details>+<summary>Prompts used</summary>++1. **Dragon:** *"TODO"*+2. **Wizard:** *"TODO"*+3. **Hut:** *"TODO"*+4. **Telescope:** *"TODO"*+5. **Robot:** *"TODO"*++</details>+++### Showcase 2: Text-to-3D (Qwen)++Text-to-3D via TRELLIS.2 with Qwen Image Edit texturing — well-suited for stylized objects and crisp details.++| | | | |+| :------: | :------: | :------: | :------: |+| <img src="docs/img/trellis2/qwen_barrel.gif" alt="Barrel" width="200"> | <img src="docs/img/trellis2/qwen_chest.gif" alt="Chest" width="200"> | <img src="docs/img/trellis2/qwen_crate.gif" alt="Crate" width="200"> | <img src="docs/img/trellis2/qwen_obelisk.gif" alt="Obelisk" width="200"> |+| *Barrel* | *Chest* | *Crate* | *Obelisk* |+| <img src="docs/img/trellis2/qwen_robot.gif" alt="Robot" width="200"> | | | |+| *Robot* | | | |++<details>+<summary>Prompts used</summary>++1. **Barrel:** *"TODO"*+2. **Chest:** *"TODO"*+3. **Crate:** *"TODO"*+4. **Obelisk:** *"TODO"*+5. **Robot:** *"TODO"*++</details>+++### Showcase 3: PBR Comparison++PBR material maps (roughness, metallic, normal) can be generated via Marigold decomposition. Each pair shows the same object without and with PBR materials.++| Non-PBR | PBR | Non-PBR | PBR |+| :------: | :------: | :------: | :------: |+| <img src="docs/img/trellis2/qwen_house.gif" alt="House (non-PBR)" width="200"> | <img src="docs/img/trellis2/qwen_house_pbr.gif" alt="House (PBR)" width="200"> | <img src="docs/img/trellis2/sdxl_wizard.gif" alt="Wizard (non-PBR)" width="200"> | <img src="docs/img/trellis2/sdxl_wizard_pbr.gif" alt="Wizard (PBR)" width="200"> |+| <small>Qwen — House</small> | <small>Qwen — House (PBR)</small> | <small>SDXL — Wizard</small> | <small>SDXL — Wizard (PBR)</small> |+| <img src="docs/img/trellis2/qwen_chest.gif" alt="Chest (non-PBR)" width="200"> | <img src="docs/img/trellis2/qwen_chest_pbr.gif" alt="Chest (PBR)" width="200"> | <img src="docs/img/trellis2/qwen_obelisk.gif" alt="Obelisk (non-PBR)" width="200"> | <img src="docs/img/trellis2/qwen_obelisk_pbr.gif" alt="Obelisk (PBR)" width="200"> |+| <small>Qwen — Chest</small> | <small>Qwen — Chest (PBR)</small> | <small>Qwen — Obelisk</small> | <small>Qwen — Obelisk (PBR)</small> |+++### Showcase 4: PBR Gallery++A selection of assets with PBR materials enabled, demonstrating realistic surface response under varying lighting.++| | | | |+| :------: | :------: | :------: | :------: |+| <img src="docs/img/trellis2/sdxl_pot_of_gold_pbr.gif" alt="Pot of gold (PBR)" width="200"> | <img src="docs/img/trellis2/sdxl_vending_machine_pbr.gif" alt="Vending machine (PBR)" width="200"> | <img src="docs/img/trellis2/qwen_astrolabe_pbr.gif" alt="Astrolabe (PBR)" width="200"> | <img src="docs/img/trellis2/qwen_tree_stump_pbr.gif" alt="Tree stump (PBR)" width="200"> |+| *Pot of Gold* | *Vending Machine* | *Astrolabe* | *Tree Stump* |+| <img src="docs/img/trellis2/qwen_rabbit_pbr.gif" alt="Rabbit (PBR)" width="200"> | <img src="docs/img/trellis2/qwen_crate_pbr.gif" alt="Crate (PBR)" width="200"> | <img src="docs/img/trellis2/qwen_obelisk_pbr.gif" alt="Obelisk (PBR)" width="200"> | <img src="docs/img/trellis2/sdxl_obelisk_pbr.gif" alt="Obelisk SDXL (PBR)" width="200"> |+| *Rabbit* | *Crate* | *Obelisk (Qwen)* | *Obelisk (SDXL)* |++<details>+<summary>Prompts used</summary>++1. **Pot of Gold:** *"TODO"*+2. **Vending Machine:** *"TODO"*+3. **Astrolabe:** *"TODO"*+4. **Tree Stump:** *"TODO"*+5. **Rabbit:** *"TODO"*+6. **Crate:** *"TODO"*+7. **Obelisk (Qwen):** *"TODO"*+8. **Obelisk (SDXL):** *"TODO"*++</details>++---++### Showcase 5: Head Stylization++Texturing an existing model using prompts and style guidance from an IPAdapter image reference.**3D Model Source:** "Brown" by ucupumar - Available at: [BlendSwap (Blend #15262)](https://www.blendswap.com/blend/15262)@@ -115,9 +202,9 @@ This showcase demonstrates how StableGen can texture the model using a standard</p>-### Showcase 2: Car Asset Texturing+### Showcase 6: Car Texturing-This showcase demonstrates how StableGen can texture a car model using different textual prompts implying various visual styles.+Texturing a car model using different prompts to achieve various visual styles.**3D Model Source:** "Pontiac GTO 67" by thecali - Available at: [BlendSwap (Blend #13575)](https://www.blendswap.com/blend/13575)@@ -127,9 +214,9 @@ This showcase demonstrates how StableGen can texture a car model using different| *Base Untextured Model* | <small>Prompt: "green car"</small> | <small>Prompt: "steampunk style car"</small> | <small>Prompt: "stealth black car"</small> |-### Showcase 3: Subway Scene Asset Texturing+### Showcase 7: Scene Texturing-This showcase demonstrates how StableGen can texture a more complex scene consisting of many mesh objects.+Texturing a complex scene consisting of many mesh objects.**3D Model Source:** "Subway Station Entrance" by argonius - Available at: [BlendSwap (Blend #19305)](https://www.blendswap.com/blend/19305)@@ -211,7 +298,7 @@ The `installer.py` script (found in this repository) automates the download and* It will download and place files into the correct subdirectories of `<YourComfyUIDirectory>`.4. **Restart ComfyUI:** If ComfyUI was running, restart it to load new custom nodes.-*(For manual dependency installation—including FLUX.1-dev and Qwen Image Edit setups—see `docs/MANUAL_INSTALLATION.md`.)*+*(For manual dependency installation-including FLUX.1-dev and Qwen Image Edit setups-see `docs/MANUAL_INSTALLATION.md`.)*### Step 3: Install StableGen Blender Plugin@@ -235,6 +322,8 @@ The `installer.py` script (found in this repository) automates the download and## 🚀 Quick Start Guide+### Texturing an Existing Model+Here’s how to get your first texture generated with StableGen:1. **Start ComfyUI Server:** Make sure it's running in the background.@@ -254,6 +343,27 @@ Here’s how to get your first texture generated with StableGen:7. **Observe:** Watch the progress in the panel and the ComfyUI console. Your object should update with the new texture! Output files will be in your specified "Output Directory".* By default, the generated texture will only be visible in the Rendered viewport shading mode (CYCLES Render Engine).+### Generating a 3D Model with TRELLIS.2++Follow these steps to generate a fully textured 3D mesh from a text prompt or reference image using the TRELLIS.2 pipeline:++1. **Prerequisites:** Make sure you have the TRELLIS.2 dependencies installed (see [Installation - Step 2](#step-2-install-dependencies-custom-nodes--ai-models---automated-recommended)) and that your hardware meets the [System Requirements](#-system-requirements).+2. **Choose a Preset:** Select and apply one of the **(MESH + TEXTURE)** labeled presets:+ * **SDXL** - best for creative, prompt-driven workflows.+ * **Qwen Image Edit** - well-suited for stylized generations, legible text, and specific details. Particularly effective for image-to-3D workflows (turning a picture into a 3D model).+ * Hover over any preset in Blender for a detailed description of what it does.+ * Alternatively, use the **TRELLIS.2 (MESH ONLY)** preset if you only need the generated mesh without automatic texturing.+3. **Select Input Mode:** Set the **`Generate from`** field to **`Prompt`** for text-to-3D, or **`Image`** to use a reference image.+4. **Provide Input:** Write a descriptive prompt or load a reference image.+5. *(Optional)* **Enable PBR:** Turn on **PBR generation** under *Advanced Parameters → Output & Material Settings* to produce physically-based material maps (roughness, metallic, normal).+6. **Generate:** Click the main **Generate** button and wait for the process to complete.+7. *(Optional)* **Refine the Result:** Adjust per-camera prompts and regenerate specific views, or switch to **Local Edit** mode (a preset is available) for targeted touch-ups.++**Exporting for a Game Engine:**++8. **Bake Textures:** You will most likely need to toggle UV unwrapping (within the `Bake Textures` operator) - the `Smart UV Project` mode works well in most cases.+9. **Export:** Use the built-in export tool `Export for Game Engine` or export manually from Blender.+---## 📖 Usage & Parameters Overview@@ -266,7 +376,7 @@ These are the main operational buttons and initial setup tools, generally found* **Generate / Cancel Generation (Main Button):** This is the primary button to start the AI texture generation process for meshe objects based on your current settings. It communicates with the ComfyUI backend. While processing, the button changes to "Cancel Generation," allowing you to stop the current task. Progress bars will appear below this button during generation.* **Bake Textures:** Converts the dynamic, multi-projection material StableGen creates on your meshes into a single, standard UV-mapped image texture per object. This is essential for exporting or simplifying scenes. You can set the resolution and UV unwrapping method for the bake. This option is crucial for finalizing your AI-generated textures into a portable format.-* **Add Cameras:** Set up multiple viewpoints using one of 7 placement strategies — from simple orbit rings to geometry-aware occlusion-optimized placement with per-camera aspect ratios. Use the interactive preview to fine-tune placement before confirming.+* **Add Cameras:** Set up multiple viewpoints using one of 7 placement strategies - from simple orbit rings to geometry-aware occlusion-optimized placement with per-camera aspect ratios. Use the interactive preview to fine-tune placement before confirming.* **Collect Camera Prompts:** Cycles through all cameras in your scene, allowing you to type a specific descriptive text prompt for each viewpoint (e.g., "front view," "close-up on face"). These per-camera prompts are used in conjunction with the main prompt if `Use camera prompts` is enabled in `Viewpoint Blending Settings`.### Preset Management@@ -289,7 +399,7 @@ These are your primary controls for defining the generation:* `Generate Sequentially`: Viewpoints generate one by one, using inpainting from previous views for consistency.* `Generate Using Grid`: Combines all views into a grid for a single generation pass, with an optional refinement step.* `Refine/Restyle Texture (Img2Img)`: Uses the current texture as input for an image-to-image process.- * `Local Edit`: Selectively modify specific areas by pointing cameras at them — new texture blends over the original with feathered edges.+ * `Local Edit`: Selectively modify specific areas by pointing cameras at them - new texture blends over the original with feathered edges.* `UV Inpaint Missing Areas`: Fills untextured areas on a UV map via inpainting.* **Target Objects:** Choose whether to texture all visible mesh objects or only selected ones.@@ -332,7 +442,7 @@ StableGen organizes the generated files within the `Output Directory` specified* `depth/` *(Depth pass renders)** `canny/` *(Renders processed using Canny edge decetor)** `normal/` *(Normal pass renders)*- * `baked/` *(Textures baked onto UV maps using the standalone `Bake Textures` tool)*+ * `baked/` *(Textures baked onto UV maps using the standalone `Bake Textures` tool, exported `.glb` files from the `Export for Game Engine` tool)** `generated_baked/` *(Textures baked as part of the generation process if "Bake Textures While Generating" is enabled)** `inpaint/` *(Files related to inpainting processes, e.g., for `Sequential mode`)** `render/` *(Renders of previous state used as context for inpainting)*@@ -407,11 +517,11 @@ The TRELLIS.2 feature relies on several third-party components, each with its ow| [nvdiffrast](https://github.com/NVlabs/nvdiffrast) (NVIDIA) | NVIDIA Source Code License | ❌ **Non-commercial only** || [nvdiffrec](https://github.com/NVlabs/nvdiffrec) (NVIDIA) | NVIDIA Source Code License | ❌ **Non-commercial only** |-**Important:** The NVIDIA libraries (`nvdiffrast` and `nvdiffrec`) are only used when the TRELLIS.2 **Texture Mode** is set to **"Native (TRELLIS.2)"** — specifically for UV rasterization and PBR texture baking. Their license restricts usage to *"research or evaluation purposes only and not for any direct or indirect monetary gain"* (Section 3.3). Only NVIDIA and its affiliates may use these libraries commercially.+**Important:** The NVIDIA libraries (`nvdiffrast` and `nvdiffrec`) are only used when the TRELLIS.2 **Texture Mode** is set to **"Native (TRELLIS.2)"** - specifically for UV rasterization and PBR texture baking. Their license restricts usage to *"research or evaluation purposes only and not for any direct or indirect monetary gain"* (Section 3.3). Only NVIDIA and its affiliates may use these libraries commercially.-**All other TRELLIS.2 modes do not introduce NVIDIA licensing restrictions:**-* **Shape-only mode ("None")** — does not use nvdiffrast/nvdiffrec. All other pipeline components are permissively licensed (MIT/Apache 2.0 + DINOv3 License).-* **Projection-based texture modes ("SDXL", "Flux 1", "Qwen Image Edit")** — does not use nvdiffrast/nvdiffrec. The licensing terms of the selected diffusion model apply as usual (e.g., FLUX.1-dev has its own license terms separate from the TRELLIS.2 pipeline).+**All other TRELLIS.2 modes do not introduce licensing restrictions:**+* **Shape-only mode ("None")** - does not use nvdiffrast/nvdiffrec. All other pipeline components are permissively licensed (MIT/Apache 2.0 + DINOv3 License).+* **Projection-based texture modes ("SDXL", "Flux 1", "Qwen Image Edit")** - does not use nvdiffrast/nvdiffrec. The licensing terms of the selected diffusion model apply as usual (e.g., FLUX.1-dev has its own license terms separate from the TRELLIS.2 pipeline).If you require commercial use of the "Native (TRELLIS.2)" texture mode, consider contacting NVIDIA regarding commercial licensing for nvdiffrast/nvdiffrec.@@ -455,13 +565,11 @@ The open spirit of the AI and open-source communities is what makes projects lik## 💡 List of planned featuresHere are some features we plan to implement in the future (in no particular order):-* **Advanced IPAdapter support:** Support for custom IPAdapter models, support for advanced IPAdapter parameters.* **Upscaling:** Support for upscaling generated textures.* **Custom VAE, CLIP model selection:** Ability to select custom VAE and CLIP models in addition to custom ControlNet and LoRA models.* **Refine mode improvements:** Features like brush based inpainting.-* **Z-Image support** (and eventual Z-Image editing model support)-* **Mesh generation:** Integration of mesh generation capabilities (TRELLIS.2 support now available as experimental feature).* **Brush-based inpainting:** Paint masks directly on the viewport for targeted local edits.+* **Better remeshing for TRELLIS.2:** Implementing more advanced remeshing techniques to improve the quality of generated meshes.If you have any suggestions, please feel free to open an issue!diff --git a/installer.py b/installer.pyindex 2a9e2a8..cc5b6d4 100644--- a/installer.py+++ b/installer.py@@ -625,16 +625,6 @@ DEPENDENCIES: Dict[str, Dict[str, Any]] = {},]},- # --- VLM Emission (Grounding DINO + SAM 2) ---- "cn_comfyui_segment_anything": {- "id": "cn_comfyui_segment_anything", "type": "node",- "name": "ComfyUI Segment Anything (GroundingDINO + SAM)",- "git_url": "https://github.com/storyicon/comfyui_segment_anything.git",- "target_dir_relative": "custom_nodes",- "repo_name": "comfyui_segment_anything",- "license": "Apache 2.0", "packages": ["vlm_emission"],- "pip_packages": ["segment_anything"],- },# --- FLUX.2 Klein 4B ---"model_flux2_klein_4b": {"id": "model_flux2_klein_4b", "type": "model",@@ -712,14 +702,7 @@ MENU_PACKAGES: Dict[str, Dict[str, Any]] = {"description_suffix": "*Installs ComfyUI_StableDelight_ll custom node + downloads the*\n"" *Stable-X/yoso-delight-v0-4-base model (~3.3GB fp16) for specular-free albedo.*",},- '11': {"name": "[VLM EMISSION] Grounding DINO + SAM Segmentation Node",- "tags": ["vlm_emission"],- "size_gb": 0.01,- "description_suffix": "*Installs comfyui_segment_anything custom node for VLM-driven emission extraction.*\n"- " *GroundingDINO_SwinB (~938MB) and sam_vit_h (~2.56GB) models are downloaded*\n"- " *automatically on first use.*",- },- '12': {"name": "[FLUX.2 KLEIN] Klein 4B FP8 + Qwen 3 Text Encoder + VAE",+ '11': {"name": "[FLUX.2 KLEIN] Klein 4B FP8 + Qwen 3 Text Encoder + VAE","tags": ["flux2_klein"],"size_gb": 12.4,"description_suffix": "*Downloads FLUX.2 Klein 4B FP8 diffusion model (~4.1GB), Qwen 3 4B text encoder bf16 (~8.0GB),*\n"diff --git a/stablegen/__init__.py b/stablegen/__init__.pyindex d0909e5..9d17200 100644--- a/stablegen/__init__.py+++ b/stablegen/__init__.py@@ -4,7 +4,7 @@ from .stablegen import StableGenPanel, ApplyPreset, SavePreset, DeletePreset, gefrom .stablegen import SG_UL_SceneQueueList, SceneQueueAdd, SceneQueueRemove, SceneQueueClear, SceneQueueMoveUp, SceneQueueMoveDown, SceneQueueOpenResult, SceneQueueProcessfrom .render_tools import BakeTextures, AddCameras, CloneCamera, MirrorCamera, ToggleCameraLabels, SwitchMaterial, ExportOrbitGIF, ExportForGameEngine, CollectCameraPrompts, CameraPromptItem, CameraOrderItem, SG_UL_CameraOrderList, SyncCameraOrder, MoveCameraOrder, ApplyCameraOrderPresetfrom .debug_tools import debug_classes as _debug_classes-from .utils import AddHDRI, ApplyModifiers, CurvesToMesh+from .utils import AddHDRI, ApplyModifiers, CurvesToMesh, sg_modal_activefrom .generator import ComfyUIGenerate, Reproject, Regenerate, MirrorReproject, Trellis2Generateimport osimport requests@@ -114,6 +114,12 @@ def _run_async(work_fn, done_fn, poll_interval=0.25, track_generation=False):def _poll():# Discard result if the server address changed after we started.if gen != _async_generation:+ # Still call done_fn so it can clean up (e.g. decrement+ # _pending_refreshes). Pass None to signal stale/no result.+ try:+ done_fn(None)+ except Exception:+ _traceback.print_exc()return None # stop polling — staleif t.is_alive():return poll_interval # keep polling@@ -518,7 +524,7 @@ class CheckServerStatus(bpy.types.Operator):# Also check that another check isn't running if using threading later# global _is_refreshing# return prefs and prefs.preferences.server_address and not _is_refreshing- return prefs and prefs.preferences.server_address # Simplified for sync check+ return prefs and prefs.preferences.server_address and not sg_modal_active(context) # Simplified for sync checkdef execute(self, context):prefs = context.preferences.addons[__package__].preferences@@ -600,7 +606,7 @@ class RefreshControlNetMappings(bpy.types.Operator):def poll(cls, context):# Can run if server address is setprefs = context.preferences.addons.get(__package__)- return prefs and prefs.preferences.server_address+ return prefs and prefs.preferences.server_address and not sg_modal_active(context)def execute(self, context):prefs = context.preferences.addons.get(__package__)@@ -1097,7 +1103,7 @@ class RefreshCheckpointList(bpy.types.Operator):@classmethoddef poll(cls, context):prefs = context.preferences.addons.get(__package__)- return prefs and prefs.preferences.server_address+ return prefs and prefs.preferences.server_address and not sg_modal_active(context)def execute(self, context):prefs = context.preferences.addons.get(__package__)@@ -1156,12 +1162,19 @@ class RefreshCheckpointList(bpy.types.Operator):_cached_checkpoint_architecture = archprint(f"Checkpoint refresh: {len(items)} {desc}(s) found.")- # Reset model_name if current selection is no longer valid+ # Reset model_name if current selection is no longer valid.+ # Prefer the sg_model_name_backup (plain StringProperty) which+ # survives file loads even when the dynamic Enum items were stale.scene = bpy.context.scene if hasattr(bpy.context, 'scene') else Noneif scene:- current = scene.model_namevalid_ids = {it[0] for it in _cached_checkpoint_list}- if current not in valid_ids:+ backup = getattr(scene, 'sg_model_name_backup', '')+ if backup and backup in valid_ids:+ # Restore from backup — this is the value that was+ # selected before the file was saved / queue snapshot.+ if scene.model_name != backup:+ scene.model_name = backup+ elif scene.model_name not in valid_ids:placeholder = next((it[0] for it in _cached_checkpoint_listif it[0].startswith("NO_") or it[0] == "NONE_FOUND"), None)if placeholder:@@ -1189,7 +1202,7 @@ class RefreshLoRAList(bpy.types.Operator):@classmethoddef poll(cls, context):prefs = context.preferences.addons.get(__package__)- return prefs and prefs.preferences.server_address+ return prefs and prefs.preferences.server_address and not sg_modal_active(context)def execute(self, context):prefs = context.preferences.addons.get(__package__)@@ -1272,6 +1285,10 @@ class AddControlNetUnit(bpy.types.Operator):update=update_parameters) # type: ignore+ @classmethod+ def poll(cls, context):+ return not sg_modal_active(context)+def invoke(self, context, event):# Always prompt for unit type and model selectionreturn context.window_manager.invoke_props_dialog(self)@@ -1316,6 +1333,10 @@ class RemoveControlNetUnit(bpy.types.Operator):update=update_parameters) # type: ignore+ @classmethod+ def poll(cls, context):+ return not sg_modal_active(context)+def invoke(self, context, event):units = context.scene.controlnet_unitsif len(units) == 1:@@ -1375,7 +1396,7 @@ class AddLoRAUnit(bpy.types.Operator):cls.poll_message_set("All available distinct LoRA models appear to have corresponding units.")return False- return True+ return not sg_modal_active(context)def execute(self, context):loras = context.scene.lora_units@@ -1434,8 +1455,9 @@ class RemoveLoRAUnit(bpy.types.Operator):def poll(cls, context):scene = context.scene# Operator can run if there are LoRA units AND the current index is valid- return len(scene.lora_units) > 0 and \- 0 <= scene.lora_units_index < len(scene.lora_units)+ return (len(scene.lora_units) > 0 and+ 0 <= scene.lora_units_index < len(scene.lora_units)+ and not sg_modal_active(context))def execute(self, context):loras = context.scene.lora_units@@ -1494,7 +1516,8 @@ def _sg_queue_save(processing=False, current_idx=0, phase='idle'):'sg_queue_gif_interpolation', 'sg_queue_gif_use_hdri','sg_queue_gif_hdri_path', 'sg_queue_gif_hdri_strength','sg_queue_gif_hdri_rotation', 'sg_queue_gif_env_mode',- 'sg_queue_gif_shadow_plane', 'sg_queue_gif_denoiser',+ 'sg_queue_gif_denoiser', 'sg_queue_gif_use_gpu',+ 'sg_queue_gif_also_no_pbr',)for key in _GIF_KEYS:if hasattr(wm, key):@@ -1771,12 +1794,17 @@ def register():],default='FIXED')- bpy.types.WindowManager.sg_queue_gif_shadow_plane = bpy.props.BoolProperty(- name="Ground Shadow", default=False- )bpy.types.WindowManager.sg_queue_gif_denoiser = bpy.props.BoolProperty(name="Denoise", default=True)+ bpy.types.WindowManager.sg_queue_gif_use_gpu = bpy.props.BoolProperty(+ name="GPU Compute", default=True,+ description="Use GPU for Cycles rendering (Blender 5.1+ only)"+ )+ bpy.types.WindowManager.sg_queue_gif_also_no_pbr = bpy.props.BoolProperty(+ name="Also Export Without PBR", default=False,+ description="After the main GIF export, disable PBR, reproject, and export a second GIF with emission-only shading (1 sample)"+ )def initial_refresh():print("StableGen: Performing initial model list refresh...")@@ -1806,10 +1834,26 @@ def register():bpy.types.Scene.comfyui_prompt = bpy.props.StringProperty(name="ComfyUI Prompt",- description="Text prompt for generation. Use || to separate generation (left) from texturing (right) prompt",+ description="Text prompt for generation (also used for texturing unless a separate texture prompt is provided)",default="gold cube",update=update_parameters)+ bpy.types.Scene.use_separate_texture_prompt = bpy.props.BoolProperty(+ name="Use Separate Texture Prompt",+ description="Enable a dedicated prompt field for the texturing pass instead of reusing the main prompt",+ default=False,+ update=update_parameters+ )+ bpy.types.Scene.texture_prompt = bpy.props.StringProperty(+ name="Texture Prompt",+ description="Prompt used for the per-camera texturing step. "+ "Avoid view-specific details here (e.g. 'logo on the front') — "+ "this prompt is applied from every camera angle. "+ "Describe side-specific features in the main prompt or in per-camera prompts instead. "+ "When empty, the main prompt is used",+ default="",+ update=update_parameters+ )bpy.types.Scene.comfyui_negative_prompt = bpy.props.StringProperty(name="ComfyUI Negative Prompt",description="Enter the negative text prompt for ComfyUI generation",@@ -1822,6 +1866,15 @@ def register():items=update_model_list,update=update_parameters)+ # Plain-string mirror of model_name so the value survives file loads+ # even when the dynamic EnumProperty items callback returns a stale+ # list (e.g. during queue processing across different architectures).+ bpy.types.Scene.sg_model_name_backup = bpy.props.StringProperty(+ name="Model Name Backup",+ description="Internal: stores the last known-good model_name identifier",+ default="",+ options={'HIDDEN'},+ )bpy.types.Scene.seed = bpy.props.IntProperty(name="Seed",description="Seed for image generation",@@ -3045,15 +3098,6 @@ def register():update=update_parameters)- bpy.types.Scene.apply_bsdf = bpy.props.BoolProperty(- name ="Apply BSDF",- description="""Apply the BSDF shader to the material- - when set to FALSE, the material will be emissive and will not be affected by the scene lighting- - when set to TRUE, the material will be affected by the scene lighting""",- default=False,- update=update_parameters- )-bpy.types.Scene.generation_mode = bpy.props.EnumProperty(name="Generation Mode",description="Controls the generation behavior",@@ -3285,8 +3329,8 @@ def register():name="Max Tokens",description="Max sparse-voxel tokens during cascade upsampling (only affects cascade modes). ""Higher = more detail but more VRAM. "- "Try 32768 or 24576 if running out of VRAM",- default=49152,+ "Increase to 49152 for maximum detail if VRAM allows",+ default=32768,min=16384,max=65536,step=4096,@@ -3337,12 +3381,6 @@ def register():default=False,update=update_parameters)- bpy.types.Scene.trellis2_low_vram = bpy.props.BoolProperty(- name="Low VRAM BG Removal",- description="Use low VRAM mode for background removal (BiRefNet)",- default=False,- update=update_parameters- )bpy.types.Scene.trellis2_bg_removal = bpy.props.EnumProperty(name="Background Removal",description="Method for removing the image background before TRELLIS.2 processing",@@ -3426,26 +3464,6 @@ def register():default=True,update=update_parameters)- bpy.types.Scene.pbr_normal_mode = bpy.props.EnumProperty(- name="Normal Mode",- description="How to apply the predicted normal map to the mesh",- items=[- ('world', "World Space",- "Camera-space normals are converted to world space. "- "Works on any geometry (no tangent frame needed). "- "Best general-purpose mode"),- ('bump', "Bump from Normal",- "Convert the normal map to a bump/height map. "- "Works on any geometry including voxel remesh, "- "but loses directional detail"),- ('tangent', "Tangent Space",- "Standard tangent-space normal map. "- "Best quality on properly UV-unwrapped meshes, "- "may show triangle artifacts on voxel-remeshed geometry"),- ],- default='world',- update=update_parameters,- )bpy.types.Scene.pbr_normal_strength = bpy.props.FloatProperty(name="Normal Strength",description="How strongly the normal map perturbs the surface shading. "@@ -3528,10 +3546,6 @@ def register():('hsv', "HSV Threshold","Isolate pixels with high saturation + high value in HSV ""space. Fast, zero model cost, best for neon/sci-fi styles"),- ('vlm_seg', "VLM Segmentation",- "Use Grounding DINO + SAM 2 to segment objects matching "- "emissive keywords (neon, screen, fire, LED …). Context-aware "- "but requires GroundingDINO + SAM2 ComfyUI nodes"),],default='residual',update=update_parameters@@ -3582,18 +3596,10 @@ def register():precision=1,update=update_parameters)- bpy.types.Scene.pbr_emission_keywords = bpy.props.StringProperty(- name="Emission Keywords",- description="(VLM Segmentation method) Comma-separated keywords "- "for Grounding DINO to search for emissive objects",- default="neon sign, LED, fire, candle flame, screen display, "- "laser, hologram, glowing crystal",- update=update_parameters- )bpy.types.Scene.pbr_emission_strength = bpy.props.FloatProperty(name="Emission Strength",description="Strength of the emission channel in the Principled BSDF",- default=5.0,+ default=2.5,min=0.0,max=100.0,step=0.5,@@ -3707,6 +3713,24 @@ def register():max=10,update=update_parameters)+ bpy.types.Scene.pbr_albedo_auto_saturation = bpy.props.BoolProperty(+ name="Correct Albedo Saturation",+ description="Automatically correct albedo saturation by comparing the PBR albedo "+ "against the original rendered image and boosting to match. "+ "Averaged across all cameras for uniform results by default. "+ "Recommended for Marigold IID (Flat Albedo) which tends to desaturate. "+ "Usually not needed for StableDelight or IID-Lighting",+ default=True,+ update=update_parameters+ )+ bpy.types.Scene.pbr_albedo_per_camera_saturation = bpy.props.BoolProperty(+ name="Per Camera",+ description="Compute saturation correction individually for each camera instead of "+ "averaging across all cameras. May produce slightly different albedo "+ "colours per view if the cameras have different lighting conditions",+ default=False,+ update=update_parameters+ )bpy.types.Scene.pbr_replace_color_with_albedo = bpy.props.BoolProperty(name="Use Albedo as Base Color",description="Replace the projected colour texture with the albedo map. "@@ -3753,6 +3777,14 @@ def register():max=10,update=update_parameters)+ bpy.types.Scene.trellis2_artifact_fail_on_exhausted = bpy.props.BoolProperty(+ name="Fail on Exhausted Retries",+ description="When enabled, the generation will fail instead of proceeding with a "+ "potentially corrupt mesh if all artifact-filter retries are exhausted. "+ "In queue mode this counts as a failure and triggers a queue-level retry",+ default=False,+ update=update_parameters+ )def unregister():"""@@ -3767,8 +3799,11 @@ def unregister():del bpy.types.Scene.use_flux_loradel bpy.types.Scene.comfyui_prompt+ del bpy.types.Scene.use_separate_texture_prompt+ del bpy.types.Scene.texture_promptdel bpy.types.Scene.comfyui_negative_promptdel bpy.types.Scene.model_name+ del bpy.types.Scene.sg_model_name_backupdel bpy.types.Scene.seeddel bpy.types.Scene.control_after_generatedel bpy.types.Scene.steps@@ -3891,13 +3926,14 @@ def unregister():'pbr_map_emission', 'pbr_emission_method','pbr_emission_threshold', 'pbr_emission_saturation_min','pbr_emission_value_min', 'pbr_emission_bloom',- 'pbr_emission_keywords', 'pbr_emission_strength',- 'pbr_normal_mode', 'pbr_normal_strength',+ 'pbr_emission_strength',+ 'pbr_normal_strength','pbr_delight_strength','pbr_use_native_resolution', 'pbr_tiling','pbr_tile_albedo', 'pbr_tile_material', 'pbr_tile_normal', 'pbr_tile_height','pbr_tile_emission', 'pbr_tile_grid', 'pbr_tile_superres','pbr_processing_resolution', 'pbr_denoise_steps', 'pbr_ensemble_size',+ 'pbr_albedo_auto_saturation', 'pbr_albedo_per_camera_saturation','pbr_replace_color_with_albedo', 'pbr_auto_lighting','pbr_model_variant', # legacy, kept for compat]@@ -3931,10 +3967,11 @@ def unregister():'trellis2_texture_size', 'trellis2_decimation', 'trellis2_remesh','trellis2_post_processing_enabled','trellis2_auto_lighting',- 'trellis2_skip_texture', 'trellis2_low_vram', 'trellis2_bg_removal', 'trellis2_background_color',+ 'trellis2_skip_texture', 'trellis2_bg_removal', 'trellis2_background_color','trellis2_fill_holes','trellis2_artifact_laplacian_sigma', 'trellis2_artifact_spike_abs_max',- 'trellis2_artifact_max_retries', 'show_trellis2_artifact_filter',+ 'trellis2_artifact_max_retries', 'trellis2_artifact_fail_on_exhausted',+ 'show_trellis2_artifact_filter',]for prop in trellis2_props:if hasattr(bpy.types.Scene, prop):@@ -3954,7 +3991,8 @@ def unregister():'sg_queue_gif_interpolation', 'sg_queue_gif_use_hdri','sg_queue_gif_hdri_path', 'sg_queue_gif_hdri_strength','sg_queue_gif_hdri_rotation', 'sg_queue_gif_env_mode',- 'sg_queue_gif_shadow_plane', 'sg_queue_gif_denoiser',+ 'sg_queue_gif_denoiser', 'sg_queue_gif_use_gpu',+ 'sg_queue_gif_also_no_pbr',):if hasattr(bpy.types.WindowManager, attr):delattr(bpy.types.WindowManager, attr)diff --git a/stablegen/generator.py b/stablegen/generator.pyindex cc9e386..ead1ff7 100644--- a/stablegen/generator.py+++ b/stablegen/generator.py@@ -634,6 +634,16 @@ class ComfyUIGenerate(bpy.types.Operator):context.scene.generation_method = 'sequential' # Force sequential for edit modelsrender = bpy.context.scene.render++ # Force 100% resolution scale so depth/guide map renders match the+ # base resolution_x × resolution_y. The Qwen workflow derives its+ # output size from the uploaded guidance map, so a non-100% percentage+ # causes it to generate at the wrong resolution.+ self._original_resolution_percentage = render.resolution_percentage+ self._original_resolution_x = render.resolution_x+ self._original_resolution_y = render.resolution_y+ render.resolution_percentage = 100+resolution_x = render.resolution_xresolution_y = render.resolution_ytotal_pixels = resolution_x * resolution_y@@ -676,6 +686,9 @@ class ComfyUIGenerate(bpy.types.Operator):context.scene.generation_status = 'idle'context.scene.sg_last_gen_error = TrueComfyUIGenerate._is_running = False+ render.resolution_percentage = self._original_resolution_percentage+ render.resolution_x = self._original_resolution_x+ render.resolution_y = self._original_resolution_yreturn {'CANCELLED'}# Sort cameras by nameself._cameras.sort(key=lambda x: x.name)@@ -723,6 +736,9 @@ class ComfyUIGenerate(bpy.types.Operator):context.scene.generation_status = 'idle'context.scene.sg_last_gen_error = TrueComfyUIGenerate._is_running = False+ render.resolution_percentage = self._original_resolution_percentage+ render.resolution_x = self._original_resolution_x+ render.resolution_y = self._original_resolution_yreturn {'CANCELLED'}# If there are curves within the scene, warn the user@@ -744,6 +760,9 @@ class ComfyUIGenerate(bpy.types.Operator):context.scene.generation_status = 'idle'context.scene.sg_last_gen_error = TrueComfyUIGenerate._is_running = False+ render.resolution_percentage = self._original_resolution_percentage+ render.resolution_x = self._original_resolution_x+ render.resolution_y = self._original_resolution_yreturn {'CANCELLED'}else: # allself._to_texture = [obj for obj in bpy.context.view_layer.objects if obj.type == 'MESH' and not obj.hide_get()]@@ -763,6 +782,9 @@ class ComfyUIGenerate(bpy.types.Operator):context.scene.generation_status = 'idle'context.scene.sg_last_gen_error = TrueComfyUIGenerate._is_running = False+ render.resolution_percentage = self._original_resolution_percentage+ render.resolution_x = self._original_resolution_x+ render.resolution_y = self._original_resolution_yreturn {'CANCELLED'}if not context.scene.overwrite_material or self._material_id == -1 or (context.scene.generation_method == 'local_edit' or (context.scene.model_architecture.startswith('qwen') and context.scene.qwen_generation_method == 'local_edit')):@@ -834,6 +856,9 @@ class ComfyUIGenerate(bpy.types.Operator):self.report({'ERROR'}, f"Cannot use UV inpainting with the material of object '{obj.name}': incompatible material. The generated material has to be active.")context.scene.generation_status = 'idle'ComfyUIGenerate._is_running = False+ render.resolution_percentage = self._original_resolution_percentage+ render.resolution_x = self._original_resolution_x+ render.resolution_y = self._original_resolution_yreturn {'CANCELLED'}# Export visibility masks for each object@@ -923,6 +948,11 @@ class ComfyUIGenerate(bpy.types.Operator):if not self._thread.is_alive():context.window_manager.event_timer_remove(self._timer)ComfyUIGenerate._is_running = False+ # Restore resolution_percentage that was forced to 100 in execute()+ if hasattr(self, '_original_resolution_percentage'):+ bpy.context.scene.render.resolution_percentage = self._original_resolution_percentage+ bpy.context.scene.render.resolution_x = self._original_resolution_x+ bpy.context.scene.render.resolution_y = self._original_resolution_y# Restore original visibility for non-selected objectsif context.scene.texture_objects == 'selected':for obj in bpy.context.view_layer.objects:@@ -1568,6 +1598,11 @@ class ComfyUIGenerate(bpy.types.Operator):existing = self._find_existing_pbr_maps(context, cam_idx)if existing:+ # Ensure raw copies exist for backward compat+ # (pre-feature files). Batch post-processing+ # runs after the loop collects all cameras.+ self._ensure_raw_copies(+ context, cam_idx, existing)self._pbr_maps[cam_idx] = existingprint(f"[StableGen] Reusing existing PBR maps "f"for camera {cam_idx}")@@ -1578,6 +1613,12 @@ class ComfyUIGenerate(bpy.types.Operator):if camera_images:self._run_pbr_decomposition_batched(context, camera_images)+ # ── Batch post-processing (albedo sat/contrast, roughness scale) ──+ # Runs after ALL cameras have raw maps saved, so auto-saturation+ # can average across all cameras uniformly.+ if self._pbr_maps:+ self._apply_albedo_postprocessing_batch(context)+def image_project_callback():if context.scene.generation_method == 'sequential':# In sequential mode, projection happened per-camera inside the loop.@@ -2047,6 +2088,236 @@ class ComfyUIGenerate(bpy.types.Operator):print(f" Warning: camera→world normal conversion failed: {err}")return None+ # ── PBR post-processing helpers ───────────────────────────────────++ @staticmethod+ def _compute_image_mean_saturation(image_path_or_bytes, sample_limit=100_000):+ """Compute the mean HSV saturation for a single image.++ *image_path_or_bytes* can be a file path (str) or raw PNG bytes.+ Skips very dark pixels whose saturation is unreliable.+ Returns a float in [0, 1], or 0.0 on error.+ """+ import colorsys+ from PIL import Image++ try:+ if isinstance(image_path_or_bytes, (str, bytes)) and isinstance(image_path_or_bytes, str):+ img = Image.open(image_path_or_bytes).convert('RGB')+ else:+ img = Image.open(io.BytesIO(image_path_or_bytes)).convert('RGB')++ pixels = list(img.getdata())+ step = max(1, len(pixels) // sample_limit)+ total_s, n = 0.0, 0+ for i in range(0, len(pixels), step):+ r, g, b = pixels[i]+ _, s, v = colorsys.rgb_to_hsv(r / 255, g / 255, b / 255)+ if v > 0.05:+ total_s += s+ n += 1+ return total_s / n if n > 0 else 0.0+ except Exception as err:+ print(f" Warning: saturation computation failed: {err}")+ return 0.0++ def _compute_avg_auto_saturation(self, context):+ """Compute the average auto-saturation ratio across ALL cameras.++ Compares each camera's original rendered image against its raw+ PBR albedo, computes per-camera ratios, then averages them.+ This ensures uniform albedo saturation regardless of per-camera+ lighting differences.++ Returns a float multiplier (1.0 = no correction needed).+ """+ ratios = []+ num_cameras = len(self._cameras) if hasattr(self, '_cameras') else 0++ for cam_idx in range(num_cameras):+ original_path = get_file_path(+ context, "generated", camera_id=cam_idx,+ material_id=self._material_id)+ raw_path = get_file_path(+ context, "pbr", subtype="raw_albedo",+ camera_id=cam_idx,+ material_id=self._material_id)++ if not os.path.exists(original_path) or not os.path.exists(raw_path):+ continue++ s_orig = self._compute_image_mean_saturation(original_path)+ s_albedo = self._compute_image_mean_saturation(raw_path)++ if s_albedo < 0.01:+ continue # skip grayscale albedos++ ratio = s_orig / s_albedo+ ratio = max(0.5, min(ratio, 3.0))+ ratios.append(ratio)++ if not ratios:+ return 1.0++ avg = sum(ratios) / len(ratios)+ print(f"[StableGen] Auto saturation: per-camera ratios = "+ f"{[f'{r:.2f}' for r in ratios]}, average = {avg:.2f}")+ return avg++ def _apply_albedo_postprocessing_batch(self, context):+ """Apply auto-saturation correction to all camera albedos.++ Must be called AFTER decomposition saves raw files for all cameras.+ By default computes the averaged saturation ratio across all cameras+ for uniform correction. With per-camera mode, each camera gets its+ own individual ratio.++ When correction is disabled, restores raw (unprocessed) albedos+ so that a previously corrected active file doesn't persist.+ """+ auto_on = getattr(context.scene, 'pbr_albedo_auto_saturation', True)++ if not auto_on:+ self._restore_raw_albedos(context)+ return++ per_camera = getattr(context.scene, 'pbr_albedo_per_camera_saturation', False)++ if per_camera:+ # Per-camera mode: individual ratio per camera+ self._apply_per_camera_saturation(context)+ else:+ # Uniform mode: averaged ratio across all cameras+ effective_sat = self._compute_avg_auto_saturation(context)++ if effective_sat == 1.0:+ print("[StableGen] Albedo saturation correction: no-op (ratio ≈ 1.0)")+ self._restore_raw_albedos(context)+ return++ for cam_idx in sorted(self._pbr_maps.keys()):+ cam_maps = self._pbr_maps[cam_idx]+ if 'albedo' not in cam_maps:+ continue++ raw_path = get_file_path(+ context, "pbr", subtype="raw_albedo",+ camera_id=cam_idx,+ material_id=self._material_id)+ if not os.path.exists(raw_path):+ continue++ with open(raw_path, 'rb') as f:+ raw_bytes = f.read()+ processed = self._postprocess_albedo(raw_bytes, effective_sat)+ with open(cam_maps['albedo'], 'wb') as f:+ f.write(processed)+ print(f" Corrected albedo saturation cam {cam_idx} "+ f"(uniform ×{effective_sat:.2f})")++ def _apply_per_camera_saturation(self, context):+ """Apply individual saturation correction for each camera."""+ for cam_idx in sorted(self._pbr_maps.keys()):+ cam_maps = self._pbr_maps[cam_idx]+ if 'albedo' not in cam_maps:+ continue++ raw_path = get_file_path(+ context, "pbr", subtype="raw_albedo",+ camera_id=cam_idx,+ material_id=self._material_id)+ original_path = get_file_path(+ context, "generated", camera_id=cam_idx,+ material_id=self._material_id)++ if not os.path.exists(raw_path) or not os.path.exists(original_path):+ continue++ s_orig = self._compute_image_mean_saturation(original_path)+ s_albedo = self._compute_image_mean_saturation(raw_path)++ if s_albedo < 0.01:+ continue++ ratio = max(0.5, min(s_orig / s_albedo, 3.0))++ with open(raw_path, 'rb') as f:+ raw_bytes = f.read()+ processed = self._postprocess_albedo(raw_bytes, ratio)+ with open(cam_maps['albedo'], 'wb') as f:+ f.write(processed)+ print(f" Corrected albedo saturation cam {cam_idx} "+ f"(per-camera ×{ratio:.2f})")++ def _restore_raw_albedos(self, context):+ """Copy raw (unprocessed) albedos back to active paths.++ Ensures that disabling saturation correction and reprojecting+ uses the original decomposition output, not a previously corrected file.+ """+ import shutil+ for cam_idx in sorted(self._pbr_maps.keys()):+ cam_maps = self._pbr_maps[cam_idx]+ if 'albedo' not in cam_maps:+ continue+ raw_path = get_file_path(+ context, "pbr", subtype="raw_albedo",+ camera_id=cam_idx,+ material_id=self._material_id)+ if os.path.exists(raw_path) and os.path.exists(cam_maps['albedo']):+ shutil.copy2(raw_path, cam_maps['albedo'])+ print(f" Restored raw albedo for camera {cam_idx}")++ @staticmethod+ def _postprocess_albedo(image_bytes, saturation=1.0):+ """Apply saturation adjustment to albedo PNG bytes.++ Returns post-processed PNG bytes, or the original bytes unchanged+ if saturation is 1.0 (no-op).+ """+ if saturation == 1.0:+ return image_bytes+ from PIL import Image, ImageEnhance+ img = Image.open(io.BytesIO(image_bytes)).convert('RGB')+ img = ImageEnhance.Color(img).enhance(saturation)+ buf = io.BytesIO()+ img.save(buf, format='PNG')+ return buf.getvalue()++ def _save_raw_copy(self, context, subtype, camera_id, data_bytes):+ """Save an unprocessed (raw) copy of a PBR map for later re-processing."""+ raw_path = get_file_path(+ context, "pbr", subtype=f"raw_{subtype}",+ camera_id=camera_id,+ material_id=self._material_id)+ try:+ with open(raw_path, 'wb') as f:+ f.write(data_bytes)+ except Exception as err:+ print(f" Warning: failed to save raw {subtype}: {err}")++ def _ensure_raw_copies(self, context, camera_id, existing_maps):+ """Ensure a raw albedo copy exists for backward compatibility.++ For files generated before the saturation correction feature, the+ active albedo IS the unprocessed original. Copy it to the raw slot+ so that the batch saturation correction can read from it.+ """+ if 'albedo' not in existing_maps:+ return+ raw_path = get_file_path(+ context, "pbr", subtype="raw_albedo",+ camera_id=camera_id,+ material_id=self._material_id)+ if not os.path.exists(raw_path) and os.path.exists(existing_maps['albedo']):+ try:+ import shutil+ shutil.copy2(existing_maps['albedo'], raw_path)+ print(f" Created missing raw albedo from active file "+ f"(camera {camera_id})")+ except Exception as err:+ print(f" Warning: could not create raw albedo copy: {err}")+def _run_pbr_decomposition_batched(self, context, camera_images):"""Run PBR decomposition **model-first** across all cameras.@@ -2126,9 +2397,9 @@ class ComfyUIGenerate(bpy.types.Operator):['_lighting_albedo', '_lighting_shading', '_lighting_residual'],))- # Count extra post-processing steps (HSV / VLM emission)+ # Count extra post-processing steps (HSV emission)extra_emission_steps = 0- if want_emission and emission_method in ('hsv', 'vlm_seg'):+ if want_emission and emission_method == 'hsv':extra_emission_steps = 1total_model_steps = (len(models_to_run)@@ -2248,6 +2519,7 @@ class ComfyUIGenerate(bpy.types.Operator):camera_id=cam_idx,material_id=self._material_id)try:+ self._save_raw_copy(context, "albedo", cam_idx, result)with open(pbr_path, 'wb') as f:f.write(result)self._pbr_maps[cam_idx]['albedo'] = pbr_path@@ -2443,19 +2715,6 @@ class ComfyUIGenerate(bpy.types.Operator):self._generate_emission_hsv(context, cam_idx, image_path)self._progress = 100- elif emission_method == 'vlm_seg':- model_step += 1- self._pbr_step = model_step- self._progress = 0- print(f"[StableGen] Emission via VLM segmentation "- f"– step {model_step}/{total_model_steps}")- for ci, cam_idx in enumerate(cam_ids):- self._pbr_cam = ci- self._stage = (f"PBR: VLM emission "- f"(cam {ci+1}/{num_cams})")- image_path = camera_images[cam_idx]- self._generate_emission_vlm(context, cam_idx, image_path)-# Deactivate PBR progress trackingself._pbr_active = Falseself._pbr_step = 0@@ -2544,6 +2803,7 @@ class ComfyUIGenerate(bpy.types.Operator):context, "pbr", subtype="albedo",camera_id=camera_id,material_id=self._material_id)+ self._save_raw_copy(context, "albedo", camera_id, map_bytes)with open(albedo_path, 'wb') as f:f.write(map_bytes)self._pbr_maps[camera_id]['albedo'] = albedo_path@@ -2574,6 +2834,10 @@ class ComfyUIGenerate(bpy.types.Operator):if converted is not None:map_bytes = converted+ # ── Raw save for albedo (post-processing applied in batch) ─+ if map_name == 'albedo':+ self._save_raw_copy(context, "albedo", camera_id, map_bytes)+with open(pbr_path, 'wb') as f:f.write(map_bytes)self._pbr_maps[camera_id][map_name] = pbr_path@@ -2701,159 +2965,6 @@ class ComfyUIGenerate(bpy.types.Operator):print(f" Failed to generate emission (HSV): {err}")import traceback; traceback.print_exc()- def _generate_emission_vlm(self, context, camera_id, image_path):- """Extract emission via Grounding DINO + SAM 2 segmentation.-- Uses ComfyUI nodes to run Grounding DINO with emissive keywords,- then SAM 2 to get pixel-perfect masks, and multiplies with the- original image for a coloured emission map.- """- scene = context.scene- keywords = getattr(scene, 'pbr_emission_keywords',- 'neon sign, LED, fire, candle flame, '- 'screen display, laser, hologram, glowing crystal')- threshold = getattr(scene, 'pbr_emission_threshold', 0.3)-- server_address = context.preferences.addons[- __package__].preferences.server_address-- # Check that required nodes exist- required_nodes = [- 'GroundingDinoModelLoader (segment anything)',- 'GroundingDinoSAMSegment (segment anything)',- 'SAMModelLoader (segment anything)',- ]- for node_class in required_nodes:- try:- resp = urllib.request.urlopen(- f"http://{server_address}/object_info/{urllib.parse.quote(node_class)}",- timeout=get_timeout('api')- )- data = json.loads(resp.read())- except Exception:- print(f" VLM emission: ComfyUI node '{node_class}' not found. "- f"Skipping VLM emission. Install comfyui_segment_anything.")- self._warning = (f"VLM Emission requires comfyui_segment_anything. "- f"Node '{node_class}' not found.")- return-- # Upload image- from .generator import upload_image_to_comfyui- image_info = upload_image_to_comfyui(server_address, image_path)- if image_info is None:- print(f" Failed to upload image for VLM emission")- return-- uploaded_name = image_info.get("name", os.path.basename(image_path))-- # Build the ComfyUI workflow:- # LoadImage → GroundingDINO detect → SAM segment → mask output- prompt = {- "1": {- "class_type": "LoadImage",- "inputs": {"image": uploaded_name}- },- "2": {- "class_type": "GroundingDinoModelLoader (segment anything)",- "inputs": {- "model_name": "GroundingDINO_SwinB (938MB)"- }- },- "3": {- "class_type": "SAMModelLoader (segment anything)",- "inputs": {- "model_name": "sam_vit_h (2.56GB)"- }- },- "4": {- "class_type": "GroundingDinoSAMSegment (segment anything)",- "inputs": {- "grounding_dino_model": ["2", 0],- "sam_model": ["3", 0],- "image": ["1", 0],- "prompt": keywords,- "threshold": threshold,- }- },- # SAM output 1 is MASK; convert to IMAGE for SaveImageWebsocket- "4b": {- "class_type": "MaskToImage",- "inputs": {"mask": ["4", 1]}- },- "5": {- "class_type": "SaveImageWebsocket",- "inputs": {"images": ["4b", 0]}- }- }-- NODES = {"save_image": "5"}-- client_id = str(uuid.uuid4())- ws = self.workflow_manager._connect_to_websocket(- server_address, client_id)- if ws is None:- print(f" VLM emission: WebSocket connection failed")- return-- try:- images = self.workflow_manager._execute_prompt_and_get_images(- ws, prompt, client_id, server_address, NODES- )- finally:- try:- ws.close()- except Exception:- pass-- if (images is None or not isinstance(images, dict)- or "5" not in images or not images["5"]):- print(f" VLM emission: No mask output received. "- f"The scene may not contain recognisable emissive objects.")- # Fall back to a black emission map- try:- orig = Image.open(image_path).convert('RGB')- black = Image.new('RGB', orig.size, (0, 0, 0))- emission_path = get_file_path(- context, "pbr", subtype="emission",- camera_id=camera_id,- material_id=self._material_id)- black.save(emission_path)- self._pbr_maps[camera_id]['emission'] = emission_path- print(f" Saved emission (VLM – no emissive objects) → "- f"{os.path.basename(emission_path)}")- except Exception:- pass- return-- mask_bytes = images["5"][0]- try:- mask = np.array(Image.open(io.BytesIO(mask_bytes)).convert('L')).astype(np.float32) / 255.0- original = np.array(Image.open(image_path).convert('RGB')).astype(np.float32) / 255.0-- # Resize mask if dimensions differ- if mask.shape[:2] != original.shape[:2]:- mask_pil = Image.fromarray((mask * 255).astype(np.uint8), mode='L')- mask_pil = mask_pil.resize(- (original.shape[1], original.shape[0]),- Image.Resampling.BILINEAR)- mask = np.array(mask_pil).astype(np.float32) / 255.0-- emission = original * mask[:, :, np.newaxis]-- emission_img = Image.fromarray(- np.clip(emission * 255, 0, 255).astype(np.uint8), mode='RGB')- emission_path = get_file_path(- context, "pbr", subtype="emission",- camera_id=camera_id,- material_id=self._material_id)- emission_img.save(emission_path)- self._pbr_maps[camera_id]['emission'] = emission_path- print(f" Saved emission (VLM segmentation) → "- f"{os.path.basename(emission_path)}")- except Exception as err:- print(f" Failed to generate emission (VLM): {err}")- import traceback; traceback.print_exc()-def _apply_qwen_context_cleanup(self, context, image_bytes):hue_tolerance = max(context.scene.qwen_context_cleanup_hue_tolerance, 0.0)value_adjust = context.scene.qwen_context_cleanup_value_adjustdiff --git a/stablegen/project.py b/stablegen/project.pyindex d5a660f..37bc47b 100644--- a/stablegen/project.py+++ b/stablegen/project.py@@ -929,24 +929,7 @@ def project_image(context, to_project, mat_id, stop_index=1000000):links.new(weight_nodes[0].outputs[0], compare_node.inputs[0])links.new(compare_node.outputs[0], final_mix.inputs["Fac"])- if not context.scene.apply_bsdf:- return final_mix, compare_node-- # Add a principle shader for the mixed color- should_add_principled = True- if is_local_edit:- # Final principled is at last_node's output if it is BSDF- final_principled = last_node.outputs[0].links[0].to_node- if final_principled.type == 'BSDF_PRINCIPLED':- should_add_principled = False-- if should_add_principled:- final_principled = nodes_collection.new("ShaderNodeBsdfPrincipled")- final_principled.location = (x_offset, y_offset)- final_principled.inputs["Roughness"].default_value = 1.0-- links.new(final_mix.outputs[0], final_principled.inputs[0])- return final_principled, compare_node+ return final_mix, compare_node# ── Helper: extract final weight output from a weight_node item ────def _get_weight_output(w_item, v_offset):@@ -2152,9 +2135,9 @@ def project_pbr_to_bsdf(context, to_texture, pbr_maps, material_id=None):breakif principled is None:- # No BSDF exists (apply_bsdf is probably False). Create one- # and rewire the existing colour output through it so that PBR- # channels have somewhere to connect.+ # No BSDF exists yet. Create one and rewire the existing+ # colour output through it so that PBR channels have+ # somewhere to connect.output_node = Nonefor node in nodes:if node.type == 'OUTPUT_MATERIAL':@@ -2327,80 +2310,23 @@ def project_pbr_to_bsdf(context, to_texture, pbr_maps, material_id=None):if sp_type == 'normal':normal_strength = getattr(scene, 'pbr_normal_strength', 0.5)- normal_mode = getattr(scene, 'pbr_normal_mode', 'world')-- if normal_mode == 'world':- # ── World-space Normal Map ────────────────────────- # The normal images were already converted from- # camera-space to world-space in- # _convert_normals_cam_to_world(). Using the- # WORLD space mode of the Normal Map node means- # we don't rely on tangent frames, so this works- # correctly on voxel-remeshed geometry.- normal_map_node = nodes.new("ShaderNodeNormalMap")- normal_map_node.space = 'WORLD'- normal_map_node.location = (- principled.location[0] - 300,- principled.location[1] - 400)- normal_map_node.label = "PBR Normal Map (World)"- normal_map_node.inputs["Strength"].default_value = (- normal_strength)- links.new(sp_output, normal_map_node.inputs["Color"])- if "Normal" in principled.inputs:- links.new(normal_map_node.outputs["Normal"],- principled.inputs["Normal"])-- elif normal_mode == 'bump':- # ── Bump from Normal ──────────────────────────────- # Convert the normal map to a height signal by- # extracting luminance, then drive a Bump node.- # This avoids tangent-space dependency so it works- # on voxel-remeshed geometry without triangle seams.- sep_node = nodes.new("ShaderNodeSeparateColor")- sep_node.location = (- principled.location[0] - 600,- principled.location[1] - 400)- sep_node.label = "PBR Normal → Channels"- links.new(sp_output, sep_node.inputs["Color"])-- bump_node = nodes.new("ShaderNodeBump")- bump_node.location = (- principled.location[0] - 300,- principled.location[1] - 400)- bump_node.label = "PBR Bump from Normal"- bump_node.inputs["Strength"].default_value = normal_strength- bump_node.inputs["Distance"].default_value = 0.02- # Use the Blue channel (Z/facing component) inverted- # as a height proxy: flat areas → high blue → low- # height; angled areas → low blue → high height.- invert_z = nodes.new("ShaderNodeMath")- invert_z.operation = 'SUBTRACT'- invert_z.inputs[0].default_value = 1.0- invert_z.location = (- principled.location[0] - 450,- principled.location[1] - 400)- invert_z.label = "Invert Z"- links.new(sep_node.outputs["Blue"],- invert_z.inputs[1])- links.new(invert_z.outputs["Value"],- bump_node.inputs["Height"])-- if "Normal" in principled.inputs:- links.new(bump_node.outputs["Normal"],- principled.inputs["Normal"])- else:- # ── Tangent-space Normal Map ──────────────────────- normal_map_node = nodes.new("ShaderNodeNormalMap")- normal_map_node.location = (- principled.location[0] - 300,- principled.location[1] - 400)- normal_map_node.label = "PBR Normal Map"- normal_map_node.inputs["Strength"].default_value = (- normal_strength)- links.new(sp_output, normal_map_node.inputs["Color"])- if "Normal" in principled.inputs:- links.new(normal_map_node.outputs["Normal"],- principled.inputs["Normal"])++ # World-space Normal Map — camera-space normals were+ # already converted in _convert_normals_cam_to_world().+ # Using WORLD space avoids tangent-frame dependency,+ # so it works on voxel-remeshed geometry.+ normal_map_node = nodes.new("ShaderNodeNormalMap")+ normal_map_node.space = 'WORLD'+ normal_map_node.location = (+ principled.location[0] - 300,+ principled.location[1] - 400)+ normal_map_node.label = "PBR Normal Map (World)"+ normal_map_node.inputs["Strength"].default_value = (+ normal_strength)+ links.new(sp_output, normal_map_node.inputs["Color"])+ if "Normal" in principled.inputs:+ links.new(normal_map_node.outputs["Normal"],+ principled.inputs["Normal"])elif sp_type == 'displacement':# ── Displacement / Height Map ─────────────────────diff --git a/stablegen/render_tools.py b/stablegen/render_tools.pyindex bd58146..96befc4 100644--- a/stablegen/render_tools.py+++ b/stablegen/render_tools.py@@ -8,7 +8,7 @@ import timeimport mathutilsfrom mathutils import Matriximport blf-from .utils import get_file_path, get_dir_path, remove_empty_dirs, get_compositor_node_tree, configure_output_node_paths, get_eevee_engine_id+from .utils import get_file_path, get_dir_path, remove_empty_dirs, get_compositor_node_tree, configure_output_node_paths, get_eevee_engine_id, sg_modal_activeimport gpufrom gpu_extras.batch import batch_for_shaderimport tempfile@@ -2383,8 +2383,12 @@ def _sg_draw_crop_overlays():shader.uniform_float("color", (_oc[0], _oc[1], _oc[2], 0.9))gpu.state.line_width_set(2.0)gpu.state.blend_set('ALPHA')+ gpu.state.depth_test_set('LESS_EQUAL')+ gpu.state.depth_mask_set(False)batch.draw(shader)+ gpu.state.depth_test_set('NONE')+ gpu.state.depth_mask_set(True)gpu.state.blend_set('NONE')gpu.state.line_width_set(1.0)@@ -4004,7 +4008,7 @@ class MirrorCamera(bpy.types.Operator):@classmethoddef poll(cls, context):- return context.scene.camera is not None+ return context.scene.camera is not None and not sg_modal_active(context)def invoke(self, context, event):return context.window_manager.invoke_props_dialog(self)@@ -4059,7 +4063,7 @@ class ToggleCameraLabels(bpy.types.Operator):@classmethoddef poll(cls, context):- return any(o.type == 'CAMERA' for o in context.scene.objects)+ return any(o.type == 'CAMERA' for o in context.scene.objects) and not sg_modal_active(context)def execute(self, context):global _sg_label_draw_handle@@ -4122,6 +4126,10 @@ class SyncCameraOrder(bpy.types.Operator):bl_label = "Sync Camera Order"bl_options = {'REGISTER', 'UNDO'}+ @classmethod+ def poll(cls, context):+ return not sg_modal_active(context)+def execute(self, context):scene = context.scenecameras = sorted(@@ -4147,6 +4155,10 @@ class MoveCameraOrder(bpy.types.Operator):name="Direction") # type: ignore+ @classmethod+ def poll(cls, context):+ return not sg_modal_active(context)+def execute(self, context):scene = context.sceneidx = scene.sg_camera_order_index@@ -4320,10 +4332,11 @@ class ApplyCameraOrderPreset(bpy.types.Operator):class CollectCameraPrompts(bpy.types.Operator):- """Cycles through cameras and prompts for a viewpoint description for each. This can help when specific perspectives fail to generate correctly.+ """Collect viewpoint description prompts for the selected cameras (or all cameras if none selected).- These prompts will be appended before the main prompt for each camera generation.- Applicable for separate, sequential and refine (also within grid mode) modes.+ - Select one or more cameras in the viewport to only prompt for those, or run with no selection to cycle through all cameras.- Examples: 'front view', 'close-up on face', 'from above'."""bl_idname = "object.collect_camera_prompts"bl_label = "Collect Camera View Prompts"@@ -4355,8 +4368,18 @@ class CollectCameraPrompts(bpy.types.Operator):return any(obj.type == 'CAMERA' for obj in context.scene.objects)def invoke(self, context, event):- # Get all camera objects and sort them by name- self._cameras = sorted([obj for obj in context.scene.objects if obj.type == 'CAMERA'], key=lambda x: x.name)+ # Use selected cameras if any are selected, otherwise fall back to all+ selected_cams = sorted(+ [obj for obj in context.selected_objects if obj.type == 'CAMERA'],+ key=lambda x: x.name,+ )+ if selected_cams:+ self._cameras = selected_cams+ else:+ self._cameras = sorted(+ [obj for obj in context.scene.objects if obj.type == 'CAMERA'],+ key=lambda x: x.name,+ )if not self._cameras:self.report({'ERROR'}, "No cameras found in the scene.")@@ -4424,6 +4447,10 @@ class SwitchMaterial(bpy.types.Operator):min=0) # type: ignore+ @classmethod+ def poll(cls, context):+ return not sg_modal_active(context)+def invoke(self, context, event):return context.window_manager.invoke_props_dialog(self)@@ -5472,15 +5499,10 @@ class BakeTextures(bpy.types.Operator):tex_base.label = "BaseColor"links.new(uv_node.outputs["UV"], tex_base.inputs["Vector"])- if context.scene.apply_bsdf or self.bake_pbr:- links.new(tex_base.outputs["Color"],- principled.inputs["Base Color"])- links.new(principled.outputs[0],- output_node.inputs["Surface"])- else:- links.new(tex_base.outputs["Color"],- output_node.inputs["Surface"])- return+ links.new(tex_base.outputs["Color"],+ principled.inputs["Base Color"])+ links.new(principled.outputs[0],+ output_node.inputs["Surface"])else:links.new(principled.outputs[0],output_node.inputs["Surface"])@@ -5698,13 +5720,6 @@ class ExportOrbitGIF(bpy.types.Operator):default='FIXED') # type: ignore- use_shadow_plane: bpy.props.BoolProperty(- name="Ground Shadow",- description="Add a temporary shadow-catcher plane beneath the "- "model for a grounded appearance (Cycles only)",- default=False- ) # type: ignore-use_denoiser: bpy.props.BoolProperty(name="Denoise",description="Enable Cycles denoising for cleaner specular "@@ -5712,6 +5727,20 @@ class ExportOrbitGIF(bpy.types.Operator):default=True) # type: ignore+ use_gpu: bpy.props.BoolProperty(+ name="GPU Compute",+ description="Use GPU for Cycles rendering (Blender 5.1+ only). "+ "Disable to force CPU rendering",+ default=True+ ) # type: ignore++ filename_suffix: bpy.props.StringProperty(+ name="Filename Suffix",+ description="Internal: appended to the output filename (e.g. '_no_pbr')",+ default="",+ options={'HIDDEN'}+ ) # type: ignore+_timer = None_rendering = False_cancelled = False # Added flag to track cancellation@@ -5719,7 +5748,6 @@ class ExportOrbitGIF(bpy.types.Operator):_handle_cancel = None_initial_settings = {}_temp_empty = None # Added for the pivot empty- _temp_shadow_plane = None # Added for shadow catcher_env_mapping_node = None # For HDRI rotation animation_original_world_data = None # Store entire world setup for restore_output_path = "" # Internal variable to store the final GIF path@@ -5728,6 +5756,7 @@ class ExportOrbitGIF(bpy.types.Operator):_frame_paths = [] # List to store paths of rendered frames_original_camera_parent = None # Store original camera parent_original_camera_matrix = None # Store original camera matrix+ _original_compute_device_type = None # For GPU/CPU toggle restore@classmethoddef poll(cls, context):@@ -5739,7 +5768,10 @@ class ExportOrbitGIF(bpy.types.Operator):except ImportError:return False# Check for active object (Mesh or Empty) and active camera- return context.active_object is not None and context.active_object.type in {'MESH', 'EMPTY'} and context.scene.camera is not None+ return (context.active_object is not None+ and context.active_object.type in {'MESH', 'EMPTY'}+ and context.scene.camera is not None+ and not sg_modal_active(context))def invoke(self, context, event):# Check dependency again in invoke to provide feedback@@ -5764,8 +5796,9 @@ class ExportOrbitGIF(bpy.types.Operator):try:revision_dir = get_dir_path(context, "revision")os.makedirs(revision_dir, exist_ok=True)- self._output_path = os.path.join(revision_dir, "orbit.gif")- self._output_path_mp4 = os.path.join(revision_dir, "orbit.mp4") # MP4 path+ suffix = self.filename_suffix or ""+ self._output_path = os.path.join(revision_dir, f"orbit{suffix}.gif")+ self._output_path_mp4 = os.path.join(revision_dir, f"orbit{suffix}.mp4") # MP4 pathexcept Exception as e:self.report({'ERROR'}, f"Could not determine output directory: {e}")return {'CANCELLED'}@@ -5953,48 +5986,6 @@ class ExportOrbitGIF(bpy.types.Operator):print("[StableGen] Orbit GIF: using existing HDRI world"f" (rotation offset={math.degrees(self.hdri_rotation):.0f}°)")- def _setup_shadow_plane(self, context):- """Create a temporary shadow-catcher plane beneath the model."""- obj = context.active_object- if not obj:- return-- # Determine the bottom of the object's bounding box in world space- bbox_corners = [obj.matrix_world @ mathutils.Vector(c)- for c in obj.bound_box]- min_z = min(c.z for c in bbox_corners)- center_x = sum(c.x for c in bbox_corners) / 8- center_y = sum(c.y for c in bbox_corners) / 8-- # Determine a reasonable plane size (2× the bounding box extent)- extent_x = max(c.x for c in bbox_corners) - min(c.x for c in bbox_corners)- extent_y = max(c.y for c in bbox_corners) - min(c.y for c in bbox_corners)- plane_size = max(extent_x, extent_y) * 3-- bpy.ops.mesh.primitive_plane_add(- size=plane_size,- location=(center_x, center_y, min_z))- plane = context.active_object- plane.name = "SG_ShadowPlane"-- # Set as shadow catcher (Cycles)- plane.is_shadow_catcher = True-- # Create a holdout material so the plane is invisible except- # for the shadow it catches.- mat = bpy.data.materials.new("SG_ShadowCatcher")- mat.use_nodes = True- mat.node_tree.nodes.clear()- output = mat.node_tree.nodes.new("ShaderNodeOutputMaterial")- holdout = mat.node_tree.nodes.new("ShaderNodeHoldout")- holdout.location = (-200, 0)- mat.node_tree.links.new(holdout.outputs[0],- output.inputs["Surface"])- plane.data.materials.append(mat)-- self._temp_shadow_plane = plane- print("[StableGen] Orbit GIF: shadow catcher plane added")-def _animate_hdri_rotation(self, context, total_frames):"""Animate the HDRI Mapping node rotation for counter-rotateor environment-only modes."""@@ -6117,8 +6108,9 @@ class ExportOrbitGIF(bpy.types.Operator):try:revision_dir = get_dir_path(context, "revision")os.makedirs(revision_dir, exist_ok=True)- self._output_path = os.path.join(revision_dir, "orbit.gif")- self._output_path_mp4 = os.path.join(revision_dir, "orbit.mp4")+ suffix = self.filename_suffix or ""+ self._output_path = os.path.join(revision_dir, f"orbit{suffix}.gif")+ self._output_path_mp4 = os.path.join(revision_dir, f"orbit{suffix}.mp4")except Exception as e:self.report({'ERROR'}, f"Could not determine output directory: {e}")return {'CANCELLED'}@@ -6137,6 +6129,8 @@ class ExportOrbitGIF(bpy.types.Operator):'filepath': render.filepath,'file_format': render.image_settings.file_format,'color_mode': render.image_settings.color_mode,+ 'resolution_x': render.resolution_x,+ 'resolution_y': render.resolution_y,'resolution_percentage': render.resolution_percentage,'use_overwrite': render.use_overwrite,'use_placeholder': render.use_placeholder,@@ -6162,6 +6156,10 @@ class ExportOrbitGIF(bpy.types.Operator):render.use_compositing = False# Apply Render Settings for PNG sequence+ # Use a fixed 1024×1024 base so GIF resolution is independent of+ # whatever the generation pipeline left in resolution_x/y.+ render.resolution_x = 1024+ render.resolution_y = 1024render.filepath = os.path.join(self._temp_dir, "frame_") # Base path for framesrender.image_settings.file_format = 'PNG' # Render as PNG sequencerender.image_settings.color_mode = 'RGBA' # Use RGBA (needed for potential alpha, even if we make background opaque)@@ -6171,6 +6169,40 @@ class ExportOrbitGIF(bpy.types.Operator):render.film_transparent = Truecycles.samples = self.samples # Set samples for rendering+ # ── GPU / CPU Compute (Blender 5.1+) ───────────────────────────+ if bpy.app.version >= (5, 1, 0) and scene.render.engine == 'CYCLES':+ try:+ cycles_prefs = bpy.context.preferences.addons['cycles'].preferences+ self._original_compute_device_type = cycles_prefs.compute_device_type+ self._initial_settings['cycles_device'] = scene.cycles.device+ if self.use_gpu:+ # Probe GPU backends, pick the first that has devices+ for backend in ('CUDA', 'OPTIX', 'HIP', 'METAL', 'ONEAPI'):+ try:+ cycles_prefs.compute_device_type = backend+ cycles_prefs.get_devices()+ gpu_devs = [d for d in cycles_prefs.devices if d.type == backend]+ if gpu_devs:+ for d in gpu_devs:+ d.use = True+ scene.cycles.device = 'GPU'+ print(f"[GIF Export] Using {backend} GPU: "+ f"{', '.join(d.name for d in gpu_devs)}")+ break+ except Exception:+ continue+ else:+ # No GPU found, fall back to CPU+ cycles_prefs.compute_device_type = 'NONE'+ scene.cycles.device = 'CPU'+ print("[GIF Export] No GPU backend found, falling back to CPU")+ else:+ cycles_prefs.compute_device_type = 'NONE'+ scene.cycles.device = 'CPU'+ except Exception as e:+ print(f"[GIF Export] Could not configure compute device: {e}")+ self._original_compute_device_type = None+# ── PBR Showcase: HDRI Environment ────────────────────────────if self.use_hdri:self._setup_hdri_environment(context)@@ -6186,10 +6218,6 @@ class ExportOrbitGIF(bpy.types.Operator):except TypeError:pass # Fallback to whatever is default- # ── PBR Showcase: Ground Shadow Plane ─────────────────────────- if self.use_shadow_plane and scene.render.engine == 'CYCLES':- self._setup_shadow_plane(context)-# Setup Animationtry:self.setup_animation(context)@@ -6299,12 +6327,16 @@ class ExportOrbitGIF(bpy.types.Operator):setattr(render.image_settings, 'color_mode', value)elif key in ('samples', 'use_denoising', 'denoiser'):setattr(cycles, key, value)+ elif key == 'cycles_device':+ cycles.device = valueelif key == 'engine': # Restore enginesetattr(render, key, value)elif key == 'film_transparent': # Restore film transparencysetattr(render, key, value)elif key == 'use_compositing': # Restore compositor pipeline staterender.use_compositing = value+ elif key in ('resolution_x', 'resolution_y'):+ setattr(render, key, value)elif hasattr(render, key):setattr(render, key, value)elif hasattr(scene, key):@@ -6316,6 +6348,15 @@ class ExportOrbitGIF(bpy.types.Operator):self._original_camera_parent = Noneself._original_camera_matrix = None+ # Restore GPU/CPU compute device type (Blender 5.1+)+ if self._original_compute_device_type is not None:+ try:+ cycles_prefs = bpy.context.preferences.addons['cycles'].preferences+ cycles_prefs.compute_device_type = self._original_compute_device_type+ except Exception as e:+ print(f"Warning: Could not restore compute device type: {e}")+ self._original_compute_device_type = None+# Remove temporary emptyif self._temp_empty:# Remove animation data@@ -6331,25 +6372,6 @@ class ExportOrbitGIF(bpy.types.Operator):bpy.data.objects.remove(self._temp_empty, do_unlink=True)self._temp_empty = None- # Remove temporary shadow-catcher plane- if self._temp_shadow_plane:- try:- # Remove the material- for mat_slot in self._temp_shadow_plane.data.materials:- if mat_slot and mat_slot.name in bpy.data.materials:- bpy.data.materials.remove(mat_slot)- # Remove the mesh data- mesh_data = self._temp_shadow_plane.data- if self._temp_shadow_plane.name in context.scene.collection.objects:- context.scene.collection.objects.unlink(self._temp_shadow_plane)- if self._temp_shadow_plane.name in bpy.data.objects:- bpy.data.objects.remove(self._temp_shadow_plane, do_unlink=True)- if mesh_data and mesh_data.name in bpy.data.meshes:- bpy.data.meshes.remove(mesh_data)- except Exception as e:- print(f"Warning: Could not remove shadow plane: {e}")- self._temp_shadow_plane = None-# Restore the world environment (remove injected mapping nodes)if self._original_world_data:try:@@ -6611,7 +6633,7 @@ class ExportForGameEngine(bpy.types.Operator):@classmethoddef poll(cls, context):addon_prefs = context.preferences.addons[__package__].preferences- return os.path.exists(addon_prefs.output_dir)+ return os.path.exists(addon_prefs.output_dir) and not sg_modal_active(context)def draw(self, context):layout = self.layoutdiff --git a/stablegen/stablegen.py b/stablegen/stablegen.pyindex a47ae48..6909b90 100644--- a/stablegen/stablegen.py+++ b/stablegen/stablegen.py@@ -5,6 +5,7 @@ import sysimport bpy # pylint: disable=import-errorimport mathutils # pylint: disable=import-errorimport math # pylint: disable=import-error+from .utils import sg_modal_activedef _is_refreshing():@@ -14,35 +15,36 @@ def _is_refreshing():# Stock presetsPRESETS = {- "DEFAULT": {"description": "Default settings for general purpose generation", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 8, "cfg": 1.5, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "hm-mvgd-hm", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.7, "sequential_factor_smooth": 0.15, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.5, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'sdxl_lightning_8step_lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]},- "MODEL IS IMPORTANT": {"description": "Same as default, but is more guided by the model", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 8, "cfg": 1.5, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.7, "sequential_factor_smooth": 0.15, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.75, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'sdxl_lightning_8step_lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]},- "CHARACTERS": {"description": "Optimized settings for character generation", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 8, "cfg": 1.5, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 80.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.7, "sequential_factor_smooth": 0.1, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "first", "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.5, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'sdxl_lightning_8step_lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]},- "CHARACTERS (ALTERNATIVE MASKING)": {"description": "Optimized for character generation. Uses alternative masking parameters to be more consistent between images, but may produce more artifacts. Try if \"Characters\" fails.", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 8, "cfg": 1.5, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 80.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.5, "sequential_factor_smooth": 0.3499999940395355, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "first", "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 10, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.5, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'sdxl_lightning_8step_lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]},- "QUICK DRAFT": {"description": "Optimized for speed", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 4, "cfg": 1.0, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "grid", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.7, "sequential_factor_smooth": 0.1, "sequential_ipadapter": True, "sequential_ipadapter_mode": "first", "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 2, "blur_mask_sigma": 1.0, "grow_mask_by": 2, "canny_threshold_low": 0, "canny_threshold_high": 80, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.5, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'Hyper-SDXL-4steps-lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]},- "UV INPAINTING": {"description": "Recommended UV Inpainting setup. It is recommended to bake texutures manually before running the generation to fine-tune unwrapping and avoid lag when generating.", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 10, "cfg": 1.5, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 80.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "uv_inpaint", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "3,0,1,2", "sequential_factor": 0.6000000238418579, "sequential_factor_smooth": 0.11000001430511475, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "first", "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 3, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "lora_units": [{'model_name': 'sdxl_lightning_8step_lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]}, # No ControlNet for UV Inpainting by default- "ARCHITECTURE": {"description": "Prioritizes only the most straight-on camera for each point. This means details generated on flat surfaces will not get blurred by getting generated differently from two or more viewpoints. Does not use visibility masking. Each picture will get generated as new, consistency depends on IPAdapter + geometry.", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 8, "cfg": 1.5, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 80.0, "weight_exponent": 10.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "separate", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": False, "sequential_custom_camera_order": "", "sequential_factor": 0.75, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "first", "ipadapter_weight_type": "style", "ipadapter_strength": 0.800000011920929, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": False, "blur_mask": True, "blur_mask_radius": 3, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.6000000238418579, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'sdxl_lightning_8step_lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]},- "QWEN PRECISE": {"description": "Precise detail when camera overlap is good. Relies on context renders plus the prompt, so sparse coverage can still introduce artifacts.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit","qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": False, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "lora_units": [{'model_name': 'Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors', 'model_strength': 1.0, 'clip_strength': 0.0}]},- "QWEN SAFE": {"description": "Safer fallback when coverage is limited. Uses the previous view (recent mode) instead of context renders to keep global look coherent, at the cost of some fine-detail persistence.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 6.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": False, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.10000000149011612, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "NONE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "lora_units": [{'model_name': 'Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors', 'model_strength': 1.0, 'clip_strength': 0.0}]},- "QWEN ALT": {"description": "Balanced option that mixes additional context renders with sequential references to smooth out coverage while keeping detail reasonable.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": False, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "recent", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.10000000149011612, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "ADDITIONAL", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "early_priority": True, "early_priority_strength": 0.5, "lora_units": [{'model_name': 'Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors', 'model_strength': 1.0, 'clip_strength': 0.0}]},- "QWEN VORONOI": {"description": "Voronoi projection mode with exponent 1000 for hard camera segmentation during generation, then resets to 15 for softer blending. Each surface point is dominated by its closest camera. Based on Qwen Precise.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 1000.0, "weight_exponent_generation_only": True, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": True, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "lora_units": [{'model_name': 'Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors', 'model_strength': 1.0, 'clip_strength': 0.0}]},- "QWEN PRECISE (NUNCHAKU)": {"description": "Precise detail using Nunchaku. Meant to be used with the Nunchaku model which has the 4-step Lightning LoRA included. Requires Nunchaku nodes.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": False, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "lora_units": []},- "QWEN SAFE (NUNCHAKU)": {"description": "Safer fallback using Nunchaku. Meant to be used with the Nunchaku model which has the 4-step Lightning LoRA included. Requires Nunchaku nodes.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 6.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": False, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.10000000149011612, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "NONE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "lora_units": []},- "QWEN ALT (NUNCHAKU)": {"description": "Balanced option using Nunchaku. Meant to be used with the Nunchaku model which has the 4-step Lightning LoRA included. Requires Nunchaku nodes.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": False, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "recent", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.10000000149011612, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "ADDITIONAL", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "early_priority": True, "early_priority_strength": 0.5, "lora_units": []},- "QWEN VORONOI (NUNCHAKU)": {"description": "Voronoi projection mode using Nunchaku. Exponent 1000 for hard camera segmentation during generation, then resets to 15 for softer blending. Requires Nunchaku nodes.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 1000.0, "weight_exponent_generation_only": True, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": True, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "lora_units": []},- "KLEIN PRECISE": {"description": "FLUX.2 Klein with precise detail when camera overlap is good. Uses depth reference images with CFGGuider (cfg=1). Best for well-covered geometry.", "control_after_generate": "fixed", "model_architecture": "flux2_klein", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": False, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour.", "qwen_custom_prompt_seq_none": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour, adopting the visual style from image 2.", "qwen_custom_prompt_seq_replace": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour. In image 2, replace all solid magenta areas with content that continues the surrounding style. Replace the background with solid gray.", "qwen_custom_prompt_seq_additional": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour. In image 2, replace all solid magenta areas with content that continues the surrounding style. Replace the background with solid gray. Image 3 represents the overall style.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "lora_units": []},- "KLEIN SAFE": {"description": "FLUX.2 Klein safer fallback when coverage is limited. Uses the previous view instead of context renders with depth references and CFGGuider (cfg=1).", "control_after_generate": "fixed", "model_architecture": "flux2_klein", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 6.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": False, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.10000000149011612, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "NONE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour.", "qwen_custom_prompt_seq_none": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour, adopting the visual style from image 2.", "qwen_custom_prompt_seq_replace": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour. In image 2, replace all solid magenta areas with content that continues the surrounding style. Replace the background with solid gray.", "qwen_custom_prompt_seq_additional": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour. In image 2, replace all solid magenta areas with content that continues the surrounding style. Replace the background with solid gray. Image 3 represents the overall style.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "lora_units": []},- "KLEIN ALT": {"description": "FLUX.2 Klein balanced option that mixes additional context renders with sequential references. Uses depth references and CFGGuider (cfg=1).", "control_after_generate": "fixed", "model_architecture": "flux2_klein", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": False, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "recent", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.10000000149011612, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "ADDITIONAL", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour.", "qwen_custom_prompt_seq_none": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour, adopting the visual style from image 2.", "qwen_custom_prompt_seq_replace": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour. In image 2, replace all solid magenta areas with content that continues the surrounding style. Replace the background with solid gray.", "qwen_custom_prompt_seq_additional": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour. In image 2, replace all solid magenta areas with content that continues the surrounding style. Replace the background with solid gray. Image 3 represents the overall style.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "early_priority": True, "early_priority_strength": 0.5, "lora_units": []},- "KLEIN VORONOI": {"description": "FLUX.2 Klein voronoi projection mode. Exponent 1000 for hard camera segmentation during generation, then resets to 15 for softer blending. Uses depth references and CFGGuider (cfg=1).", "control_after_generate": "fixed", "model_architecture": "flux2_klein", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 1000.0, "weight_exponent_generation_only": True, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": True, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour.", "qwen_custom_prompt_seq_none": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour, adopting the visual style from image 2.", "qwen_custom_prompt_seq_replace": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour. In image 2, replace all solid magenta areas with content that continues the surrounding style. Replace the background with solid gray.", "qwen_custom_prompt_seq_additional": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour. In image 2, replace all solid magenta areas with content that continues the surrounding style. Replace the background with solid gray. Image 3 represents the overall style.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "lora_units": []},- "LOCAL REFINE": {"description": "Uses the SDXL local edit mode to improve detail / refine specific areas. Use new set of cameras or a single camera pointed at the area you want to refine.", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 8, "cfg": 1.5, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 80.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "reinhard", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "auto_rescale_target_mp": 1.0, "overwrite_material": True, "generation_method": "local_edit", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.800000011920929, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.10000000149011612, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "original_render", "sequential_desaturate_factor": 0.10000000149011612, "sequential_contrast_factor": 0.10000000149011612, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_voronoi_mode": True, "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 5.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "refine_angle_ramp_active": True, "refine_angle_ramp_pos_0": 0.0, "refine_angle_ramp_pos_1": 0.05, "visibility_vignette": True, "visibility_vignette_width": 0.3, "visibility_vignette_softness": 1.0, "visibility_vignette_blur": False, "refine_feather_ramp_pos_0": 0.0, "refine_feather_ramp_pos_1": 0.6, "refine_edge_feather_projection": True, "refine_edge_feather_width": 30, "refine_edge_feather_softness": 1.0, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.5, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'sdxl_lightning_8step_lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]},- "LOCAL EDIT (QWEN)": {"description": "Uses Qwen to make targeted edits to specific areas. Point cameras at what you want to change — you can alter colors, style, rewrite text, add details, and more. Untouched areas are preserved.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "qwen_generation_method": "local_edit", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "reinhard", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "auto_rescale_target_mp": 1.0, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 1.0, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": False, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_voronoi_mode": False, "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": False, "refine_angle_ramp_active": False, "refine_angle_ramp_pos_0": 0.0, "refine_angle_ramp_pos_1": 0.05, "visibility_vignette": True, "visibility_vignette_width": 0.1, "visibility_vignette_softness": 1.0, "visibility_vignette_blur": False, "refine_feather_ramp_pos_0": 0.0, "refine_feather_ramp_pos_1": 0.6, "refine_edge_feather_projection": True, "refine_edge_feather_width": 15, "refine_edge_feather_softness": 1.0, "lora_units": [{'model_name': 'Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors', 'model_strength': 1.0, 'clip_strength': 0.0}]},- "REFINE (QWEN)": {"description": "Uses Qwen to restyle or globally modify the entire existing texture. Applies changes uniformly across all camera views — ideal for changing the overall color scheme, art style, or surface appearance.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "qwen_generation_method": "refine", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "reinhard", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "auto_rescale_target_mp": 1.0, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 1.0, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": False, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_voronoi_mode": False, "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": False, "lora_units": [{'model_name': 'Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors', 'model_strength': 1.0, 'clip_strength': 0.0}]},- "DEFAULT (MODEL + TEXTURE)": {"description": "Uses TRELLIS.2 and SDXL to generate a textured mesh. Optimized for general object generation. May not work ideally for the cases which have their specialized presets.", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 8, "cfg": 1.5, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 75.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 85.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "reinhard", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "auto_rescale_target_mp": 1.0, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.800000011920929, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.10000000149011612, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "trellis2_input", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 3, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_voronoi_mode": False, "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "NONE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 5.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "qwen_generation_method": "generate", "qwen_refine_use_prev_ref": False, "qwen_refine_use_depth": False, "qwen_timestep_zero_ref": False, "refine_angle_ramp_active": True, "refine_angle_ramp_pos_0": 0.0, "refine_angle_ramp_pos_1": 0.05000000074505806, "visibility_vignette": True, "visibility_vignette_width": 0.15000000596046448, "visibility_vignette_softness": 1.0, "visibility_vignette_blur": False, "refine_feather_ramp_pos_0": 0.0, "refine_feather_ramp_pos_1": 0.6000000238418579, "refine_edge_feather_projection": True, "refine_edge_feather_width": 30, "refine_edge_feather_softness": 1.0, "trellis2_texture_mode": "sdxl", "trellis2_initial_image_arch": "sdxl", "trellis2_camera_count": 10, "trellis2_placement_mode": "normal_weighted", "trellis2_auto_prompts": True, "trellis2_exclude_bottom": False, "trellis2_exclude_bottom_angle": 1.5533000230789185, "trellis2_auto_aspect": "per_camera", "trellis2_occlusion_mode": "none", "trellis2_consider_existing": False, "trellis2_delete_cameras": False, "trellis2_coverage_target": 0.949999988079071, "trellis2_max_auto_cameras": 12, "trellis2_fan_angle": 90.0, "trellis2_resolution": "1024_cascade", "trellis2_vram_mode": "disk_offload", "trellis2_attn_backend": "flash_attn", "trellis2_ss_guidance": 7.5, "trellis2_ss_steps": 12, "trellis2_shape_guidance": 7.5, "trellis2_shape_steps": 12, "trellis2_tex_guidance": 7.5, "trellis2_tex_steps": 12, "trellis2_max_tokens": 49152, "trellis2_texture_size": 4096, "trellis2_decimation": 1000000, "trellis2_remesh": True, "trellis2_post_processing_enabled": True, "trellis2_low_vram": False, "trellis2_bg_removal": "auto", "trellis2_background_color": "black", "trellis2_fill_holes": True, "trellis2_import_scale": 2.0, "trellis2_clamp_elevation": False, "trellis2_max_elevation": 1.2216999530792236, "trellis2_min_elevation": -1.0471975803375244, "trellis2_auto_lighting": True, "use_ipadapter": False, "sequential_ipadapter_regenerate_wo_controlnet": False, "allow_modify_existing_textures": False, "ask_object_prompts": True, "weight_exponent_mask": False, "mask_blocky": False, "architecture_mode": "trellis2", "use_camera_prompts": True, "sg_use_custom_camera_order": False, "apply_bsdf": False, "generation_mode": "standard", "texture_objects": "all", "use_flux_lora": True, "qwen_use_trellis2_style": False, "qwen_trellis2_style_initial_only": False, "trellis2_skip_texture": True, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.6000000238418579, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'sdxl_lightning_8step_lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]},- "CHARACTERS (MODEL + TEXTURE)": {"description": "Uses TRELLIS.2 and SDXL to generate a textured mesh. Optimized settings for generating characters. Will not texture bottom facing faces.", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 8, "cfg": 1.5, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 75.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 85.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "reinhard", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "auto_rescale_target_mp": 1.0, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.800000011920929, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.10000000149011612, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "trellis2_input", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_voronoi_mode": False, "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "NONE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 5.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "qwen_generation_method": "generate", "qwen_refine_use_prev_ref": False, "qwen_refine_use_depth": False, "qwen_timestep_zero_ref": False, "refine_angle_ramp_active": True, "refine_angle_ramp_pos_0": 0.0, "refine_angle_ramp_pos_1": 0.05000000074505806, "visibility_vignette": True, "visibility_vignette_width": 0.15000000596046448, "visibility_vignette_softness": 1.0, "visibility_vignette_blur": False, "refine_feather_ramp_pos_0": 0.0, "refine_feather_ramp_pos_1": 0.6000000238418579, "refine_edge_feather_projection": True, "refine_edge_feather_width": 30, "refine_edge_feather_softness": 1.0, "trellis2_texture_mode": "sdxl", "trellis2_initial_image_arch": "sdxl", "trellis2_camera_count": 8, "trellis2_placement_mode": "normal_weighted", "trellis2_auto_prompts": True, "trellis2_exclude_bottom": True, "trellis2_exclude_bottom_angle": 1.5533000230789185, "trellis2_auto_aspect": "per_camera", "trellis2_occlusion_mode": "none", "trellis2_consider_existing": False, "trellis2_delete_cameras": False, "trellis2_coverage_target": 0.949999988079071, "trellis2_max_auto_cameras": 12, "trellis2_fan_angle": 90.0, "trellis2_resolution": "1024_cascade", "trellis2_vram_mode": "disk_offload", "trellis2_attn_backend": "flash_attn", "trellis2_ss_guidance": 7.5, "trellis2_ss_steps": 12, "trellis2_shape_guidance": 7.5, "trellis2_shape_steps": 12, "trellis2_tex_guidance": 7.5, "trellis2_tex_steps": 12, "trellis2_max_tokens": 49152, "trellis2_texture_size": 4096, "trellis2_decimation": 1000000, "trellis2_remesh": True, "trellis2_post_processing_enabled": True, "trellis2_low_vram": False, "trellis2_bg_removal": "auto", "trellis2_background_color": "black", "trellis2_fill_holes": True, "trellis2_import_scale": 2.0, "trellis2_clamp_elevation": True, "trellis2_max_elevation": 1.2216999530792236, "trellis2_min_elevation": -1.0471975803375244, "trellis2_auto_lighting": True, "use_ipadapter": False, "sequential_ipadapter_regenerate_wo_controlnet": False, "allow_modify_existing_textures": False, "ask_object_prompts": True, "weight_exponent_mask": False, "mask_blocky": False, "architecture_mode": "trellis2", "use_camera_prompts": True, "sg_use_custom_camera_order": False, "apply_bsdf": False, "generation_mode": "standard", "texture_objects": "all", "use_flux_lora": True, "qwen_use_trellis2_style": False, "qwen_trellis2_style_initial_only": False, "trellis2_skip_texture": True, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.6000000238418579, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'sdxl_lightning_8step_lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]},- "ARCHITECTURE (MODEL + TEXTURE)": {"description": "Uses TRELLIS.2 and SDXL to generate a textured mesh. Optimized settings for architecture, and other models with flat walls and sharp angles.", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 8, "cfg": 1.5, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 80.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 85.0, "weight_exponent": 10.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "reinhard", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "auto_rescale_target_mp": 1.0, "overwrite_material": True, "generation_method": "separate", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.800000011920929, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": False, "sequential_custom_camera_order": "", "sequential_factor": 0.75, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "trellis2_input", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 0.800000011920929, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": False, "blur_mask": True, "blur_mask_radius": 3, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_voronoi_mode": False, "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "NONE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 5.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "qwen_generation_method": "generate", "qwen_refine_use_prev_ref": False, "qwen_refine_use_depth": False, "qwen_timestep_zero_ref": False, "refine_angle_ramp_active": True, "refine_angle_ramp_pos_0": 0.0, "refine_angle_ramp_pos_1": 0.05000000074505806, "visibility_vignette": True, "visibility_vignette_width": 0.15000000596046448, "visibility_vignette_softness": 1.0, "visibility_vignette_blur": False, "refine_feather_ramp_pos_0": 0.0, "refine_feather_ramp_pos_1": 0.6000000238418579, "refine_edge_feather_projection": True, "refine_edge_feather_width": 30, "refine_edge_feather_softness": 1.0, "trellis2_texture_mode": "sdxl", "trellis2_initial_image_arch": "sdxl", "trellis2_camera_count": 8, "trellis2_placement_mode": "normal_weighted", "trellis2_auto_prompts": True, "trellis2_exclude_bottom": True, "trellis2_exclude_bottom_angle": 1.5533000230789185, "trellis2_auto_aspect": "per_camera", "trellis2_occlusion_mode": "none", "trellis2_consider_existing": False, "trellis2_delete_cameras": False, "trellis2_coverage_target": 0.949999988079071, "trellis2_max_auto_cameras": 12, "trellis2_fan_angle": 90.0, "trellis2_resolution": "1024_cascade", "trellis2_vram_mode": "disk_offload", "trellis2_attn_backend": "flash_attn", "trellis2_ss_guidance": 7.5, "trellis2_ss_steps": 12, "trellis2_shape_guidance": 7.5, "trellis2_shape_steps": 12, "trellis2_tex_guidance": 7.5, "trellis2_tex_steps": 12, "trellis2_max_tokens": 49152, "trellis2_texture_size": 4096, "trellis2_decimation": 1000000, "trellis2_remesh": True, "trellis2_post_processing_enabled": True, "trellis2_low_vram": False, "trellis2_bg_removal": "auto", "trellis2_background_color": "black", "trellis2_fill_holes": True, "trellis2_import_scale": 2.0, "trellis2_clamp_elevation": False, "trellis2_max_elevation": 1.2216999530792236, "trellis2_min_elevation": -1.0471975803375244, "trellis2_auto_lighting": True, "use_ipadapter": False, "sequential_ipadapter_regenerate_wo_controlnet": False, "allow_modify_existing_textures": False, "ask_object_prompts": True, "weight_exponent_mask": False, "mask_blocky": False, "architecture_mode": "trellis2", "use_camera_prompts": True, "sg_use_custom_camera_order": False, "apply_bsdf": False, "generation_mode": "standard", "texture_objects": "all", "use_flux_lora": True, "qwen_use_trellis2_style": False, "qwen_trellis2_style_initial_only": False, "trellis2_skip_texture": True, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.6000000238418579, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'sdxl_lightning_8step_lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]},- "QWEN PRECISE (MODEL + TEXTURE)": {"description": "Uses TRELLIS.2 and Qwen Image Edit to generate a textured mesh. Precise detail when camera overlap is good. Relies on context renders plus the prompt.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "reinhard", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "auto_rescale_target_mp": 1.0, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.800000011920929, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_voronoi_mode": False, "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "qwen_generation_method": "generate", "qwen_refine_use_prev_ref": False, "qwen_refine_use_depth": False, "qwen_timestep_zero_ref": False, "refine_angle_ramp_active": True, "refine_angle_ramp_pos_0": 0.0, "refine_angle_ramp_pos_1": 0.05000000074505806, "visibility_vignette": True, "visibility_vignette_width": 0.15000000596046448, "visibility_vignette_softness": 1.0, "visibility_vignette_blur": False, "sg_silhouette_margin": 3, "sg_silhouette_depth": 0.05000000074505806, "sg_silhouette_rays": "4", "refine_feather_ramp_pos_0": 0.0, "refine_feather_ramp_pos_1": 0.6000000238418579, "refine_edge_feather_projection": True, "refine_edge_feather_width": 30, "refine_edge_feather_softness": 1.0, "trellis2_texture_mode": "qwen_image_edit", "trellis2_initial_image_arch": "sdxl", "trellis2_camera_count": 8, "trellis2_placement_mode": "normal_weighted", "trellis2_auto_prompts": True, "trellis2_exclude_bottom": True, "trellis2_exclude_bottom_angle": 1.5533000230789185, "trellis2_auto_aspect": "per_camera", "trellis2_occlusion_mode": "none", "trellis2_consider_existing": False, "trellis2_delete_cameras": False, "trellis2_coverage_target": 0.949999988079071, "trellis2_max_auto_cameras": 12, "trellis2_fan_angle": 90.0, "trellis2_resolution": "1024_cascade", "trellis2_vram_mode": "disk_offload", "trellis2_attn_backend": "flash_attn", "trellis2_seed": 0, "trellis2_ss_guidance": 7.5, "trellis2_ss_steps": 12, "trellis2_shape_guidance": 7.5, "trellis2_shape_steps": 12, "trellis2_tex_guidance": 7.5, "trellis2_tex_steps": 12, "trellis2_max_tokens": 32768, "trellis2_texture_size": 4096, "trellis2_decimation": 1000000, "trellis2_remesh": True, "trellis2_post_processing_enabled": True, "trellis2_low_vram": False, "trellis2_bg_removal": "auto", "trellis2_background_color": "black", "trellis2_fill_holes": True, "trellis2_import_scale": 2.0, "trellis2_clamp_elevation": True, "trellis2_max_elevation": 1.2216999530792236, "trellis2_min_elevation": -1.0471975803375244, "trellis2_auto_lighting": True, "use_ipadapter": False, "sequential_ipadapter_regenerate_wo_controlnet": False, "allow_modify_existing_textures": False, "ask_object_prompts": True, "weight_exponent_mask": False, "mask_blocky": False, "architecture_mode": "trellis2", "use_camera_prompts": True, "sg_use_custom_camera_order": False, "apply_bsdf": False, "generation_mode": "standard", "texture_objects": "all", "use_flux_lora": True, "qwen_use_trellis2_style": True, "qwen_trellis2_style_initial_only": True, "trellis2_skip_texture": True, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.6000000238418579, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors', 'model_strength': 1.0, 'clip_strength': 0.0}]},- "QWEN SAFE (MODEL + TEXTURE)": {"description": "Uses TRELLIS.2 and Qwen Image Edit to generate a textured mesh. Safer fallback when coverage is limited. Uses TRELLIS.2 input image as sequential style reference for global coherence.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 6.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "reinhard", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "auto_rescale_target_mp": 1.0, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.800000011920929, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "trellis2_input", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.10000000149011612, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_voronoi_mode": False, "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "NONE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "qwen_generation_method": "generate", "qwen_refine_use_prev_ref": False, "qwen_refine_use_depth": False, "qwen_timestep_zero_ref": False, "refine_angle_ramp_active": True, "refine_angle_ramp_pos_0": 0.0, "refine_angle_ramp_pos_1": 0.05000000074505806, "visibility_vignette": True, "visibility_vignette_width": 0.15000000596046448, "visibility_vignette_softness": 1.0, "visibility_vignette_blur": False, "sg_silhouette_margin": 3, "sg_silhouette_depth": 0.05000000074505806, "sg_silhouette_rays": "4", "refine_feather_ramp_pos_0": 0.0, "refine_feather_ramp_pos_1": 0.6000000238418579, "refine_edge_feather_projection": True, "refine_edge_feather_width": 30, "refine_edge_feather_softness": 1.0, "trellis2_texture_mode": "qwen_image_edit", "trellis2_initial_image_arch": "sdxl", "trellis2_camera_count": 8, "trellis2_placement_mode": "normal_weighted", "trellis2_auto_prompts": True, "trellis2_exclude_bottom": True, "trellis2_exclude_bottom_angle": 1.5533000230789185, "trellis2_auto_aspect": "per_camera", "trellis2_occlusion_mode": "none", "trellis2_consider_existing": False, "trellis2_delete_cameras": False, "trellis2_coverage_target": 0.949999988079071, "trellis2_max_auto_cameras": 12, "trellis2_fan_angle": 90.0, "trellis2_resolution": "1024_cascade", "trellis2_vram_mode": "disk_offload", "trellis2_attn_backend": "flash_attn", "trellis2_seed": 0, "trellis2_ss_guidance": 7.5, "trellis2_ss_steps": 12, "trellis2_shape_guidance": 7.5, "trellis2_shape_steps": 12, "trellis2_tex_guidance": 7.5, "trellis2_tex_steps": 12, "trellis2_max_tokens": 32768, "trellis2_texture_size": 4096, "trellis2_decimation": 1000000, "trellis2_remesh": True, "trellis2_post_processing_enabled": True, "trellis2_low_vram": False, "trellis2_bg_removal": "auto", "trellis2_background_color": "black", "trellis2_fill_holes": True, "trellis2_import_scale": 2.0, "trellis2_clamp_elevation": True, "trellis2_max_elevation": 1.2216999530792236, "trellis2_min_elevation": -1.0471975803375244, "trellis2_auto_lighting": True, "use_ipadapter": False, "sequential_ipadapter_regenerate_wo_controlnet": False, "allow_modify_existing_textures": False, "ask_object_prompts": True, "weight_exponent_mask": False, "mask_blocky": False, "architecture_mode": "trellis2", "use_camera_prompts": True, "sg_use_custom_camera_order": False, "apply_bsdf": False, "generation_mode": "standard", "texture_objects": "all", "use_flux_lora": True, "qwen_use_trellis2_style": True, "qwen_trellis2_style_initial_only": True, "trellis2_skip_texture": True, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.6000000238418579, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors', 'model_strength': 1.0, 'clip_strength': 0.0}]},- "QWEN ALT (MODEL + TEXTURE)": {"description": "Uses TRELLIS.2 and Qwen Image Edit to generate a textured mesh. Uses ADDITIONAL context renders with TRELLIS.2 style transfer for consistency. Good general-purpose Qwen pipeline.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "reinhard", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "auto_rescale_target_mp": 1.0, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.800000011920929, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.10000000149011612, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": True, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_voronoi_mode": False, "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "ADDITIONAL", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "qwen_generation_method": "generate", "qwen_refine_use_prev_ref": False, "qwen_refine_use_depth": False, "qwen_timestep_zero_ref": False, "refine_angle_ramp_active": True, "refine_angle_ramp_pos_0": 0.0, "refine_angle_ramp_pos_1": 0.05000000074505806, "visibility_vignette": True, "visibility_vignette_width": 0.15000000596046448, "visibility_vignette_softness": 1.0, "visibility_vignette_blur": False, "sg_silhouette_margin": 3, "sg_silhouette_depth": 0.05000000074505806, "sg_silhouette_rays": "4", "refine_feather_ramp_pos_0": 0.0, "refine_feather_ramp_pos_1": 0.6000000238418579, "refine_edge_feather_projection": True, "refine_edge_feather_width": 30, "refine_edge_feather_softness": 1.0, "trellis2_texture_mode": "qwen_image_edit", "trellis2_initial_image_arch": "sdxl", "trellis2_camera_count": 8, "trellis2_placement_mode": "normal_weighted", "trellis2_auto_prompts": True, "trellis2_exclude_bottom": True, "trellis2_exclude_bottom_angle": 1.5533000230789185, "trellis2_auto_aspect": "per_camera", "trellis2_occlusion_mode": "none", "trellis2_consider_existing": False, "trellis2_delete_cameras": False, "trellis2_coverage_target": 0.949999988079071, "trellis2_max_auto_cameras": 12, "trellis2_fan_angle": 90.0, "trellis2_resolution": "1024_cascade", "trellis2_vram_mode": "disk_offload", "trellis2_attn_backend": "flash_attn", "trellis2_seed": 0, "trellis2_ss_guidance": 7.5, "trellis2_ss_steps": 12, "trellis2_shape_guidance": 7.5, "trellis2_shape_steps": 12, "trellis2_tex_guidance": 7.5, "trellis2_tex_steps": 12, "trellis2_max_tokens": 32768, "trellis2_texture_size": 4096, "trellis2_decimation": 1000000, "trellis2_remesh": True, "trellis2_post_processing_enabled": True, "trellis2_low_vram": False, "trellis2_bg_removal": "auto", "trellis2_background_color": "black", "trellis2_fill_holes": True, "trellis2_import_scale": 2.0, "trellis2_clamp_elevation": True, "trellis2_max_elevation": 1.2216999530792236, "trellis2_min_elevation": -1.0471975803375244, "trellis2_auto_lighting": True, "use_ipadapter": False, "sequential_ipadapter_regenerate_wo_controlnet": False, "allow_modify_existing_textures": False, "ask_object_prompts": True, "weight_exponent_mask": False, "mask_blocky": False, "architecture_mode": "trellis2", "use_camera_prompts": True, "sg_use_custom_camera_order": False, "apply_bsdf": False, "generation_mode": "standard", "texture_objects": "all", "use_flux_lora": True, "qwen_use_trellis2_style": True, "qwen_trellis2_style_initial_only": True, "trellis2_skip_texture": True, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.6000000238418579, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors', 'model_strength': 1.0, 'clip_strength': 0.0}]},- "QWEN VORONOI (MODEL + TEXTURE)": {"description": "Uses TRELLIS.2 and Qwen Image Edit to generate a textured mesh. Voronoi projection with exponent 1000 for hard camera segmentation during generation, then resets to 15 for softer blending.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 1000.0, "weight_exponent_generation_only": True, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "reinhard", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "auto_rescale_target_mp": 1.0, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.800000011920929, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_voronoi_mode": True, "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "qwen_generation_method": "generate", "qwen_refine_use_prev_ref": False, "qwen_refine_use_depth": False, "qwen_timestep_zero_ref": False, "refine_angle_ramp_active": True, "refine_angle_ramp_pos_0": 0.0, "refine_angle_ramp_pos_1": 0.05000000074505806, "visibility_vignette": True, "visibility_vignette_width": 0.15000000596046448, "visibility_vignette_softness": 1.0, "visibility_vignette_blur": False, "sg_silhouette_margin": 3, "sg_silhouette_depth": 0.05000000074505806, "sg_silhouette_rays": "4", "refine_feather_ramp_pos_0": 0.0, "refine_feather_ramp_pos_1": 0.6000000238418579, "refine_edge_feather_projection": True, "refine_edge_feather_width": 30, "refine_edge_feather_softness": 1.0, "trellis2_texture_mode": "qwen_image_edit", "trellis2_initial_image_arch": "sdxl", "trellis2_camera_count": 8, "trellis2_placement_mode": "normal_weighted", "trellis2_auto_prompts": True, "trellis2_exclude_bottom": True, "trellis2_exclude_bottom_angle": 1.5533000230789185, "trellis2_auto_aspect": "per_camera", "trellis2_occlusion_mode": "none", "trellis2_consider_existing": False, "trellis2_delete_cameras": False, "trellis2_coverage_target": 0.949999988079071, "trellis2_max_auto_cameras": 12, "trellis2_fan_angle": 90.0, "trellis2_resolution": "1024_cascade", "trellis2_vram_mode": "disk_offload", "trellis2_attn_backend": "flash_attn", "trellis2_seed": 0, "trellis2_ss_guidance": 7.5, "trellis2_ss_steps": 12, "trellis2_shape_guidance": 7.5, "trellis2_shape_steps": 12, "trellis2_tex_guidance": 7.5, "trellis2_tex_steps": 12, "trellis2_max_tokens": 32768, "trellis2_texture_size": 4096, "trellis2_decimation": 1000000, "trellis2_remesh": True, "trellis2_post_processing_enabled": True, "trellis2_low_vram": False, "trellis2_bg_removal": "auto", "trellis2_background_color": "black", "trellis2_fill_holes": True, "trellis2_import_scale": 2.0, "trellis2_clamp_elevation": True, "trellis2_max_elevation": 1.2216999530792236, "trellis2_min_elevation": -1.0471975803375244, "trellis2_auto_lighting": True, "use_ipadapter": False, "sequential_ipadapter_regenerate_wo_controlnet": False, "allow_modify_existing_textures": False, "ask_object_prompts": True, "weight_exponent_mask": False, "mask_blocky": False, "architecture_mode": "trellis2", "use_camera_prompts": True, "sg_use_custom_camera_order": False, "apply_bsdf": False, "generation_mode": "standard", "texture_objects": "all", "use_flux_lora": True, "qwen_use_trellis2_style": True, "qwen_trellis2_style_initial_only": True, "trellis2_skip_texture": True, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.6000000238418579, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors', 'model_strength': 1.0, 'clip_strength': 0.0}]},+ "DEFAULT": {"description": "Default settings for general purpose generation", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 8, "cfg": 1.5, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "hm-mvgd-hm", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.7, "sequential_factor_smooth": 0.15, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "pbr_decomposition": False, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.5, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'sdxl_lightning_8step_lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]},+ "MODEL IS IMPORTANT": {"description": "Same as default, but is more guided by the model", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 8, "cfg": 1.5, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.7, "sequential_factor_smooth": 0.15, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "pbr_decomposition": False, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.75, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'sdxl_lightning_8step_lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]},+ "CHARACTERS": {"description": "Optimized settings for character generation", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 8, "cfg": 1.5, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 80.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.7, "sequential_factor_smooth": 0.1, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "first", "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "pbr_decomposition": False, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.5, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'sdxl_lightning_8step_lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]},+ "CHARACTERS (ALTERNATIVE MASKING)": {"description": "Optimized for character generation. Uses alternative masking parameters to be more consistent between images, but may produce more artifacts. Try if \"Characters\" fails.", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 8, "cfg": 1.5, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 80.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.5, "sequential_factor_smooth": 0.3499999940395355, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "first", "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 10, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "pbr_decomposition": False, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.5, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'sdxl_lightning_8step_lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]},+ "QUICK DRAFT": {"description": "Optimized for speed", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 4, "cfg": 1.0, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "grid", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.7, "sequential_factor_smooth": 0.1, "sequential_ipadapter": True, "sequential_ipadapter_mode": "first", "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 2, "blur_mask_sigma": 1.0, "grow_mask_by": 2, "canny_threshold_low": 0, "canny_threshold_high": 80, "pbr_decomposition": False, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.5, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'Hyper-SDXL-4steps-lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]},+ "UV INPAINTING": {"description": "Recommended UV Inpainting setup. It is recommended to bake texutures manually before running the generation to fine-tune unwrapping and avoid lag when generating.", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 10, "cfg": 1.5, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 80.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "uv_inpaint", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "3,0,1,2", "sequential_factor": 0.6000000238418579, "sequential_factor_smooth": 0.11000001430511475, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "first", "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 3, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "pbr_decomposition": False, "lora_units": [{'model_name': 'sdxl_lightning_8step_lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]}, # No ControlNet for UV Inpainting by default+ "ARCHITECTURE": {"description": "Prioritizes only the most straight-on camera for each point. This means details generated on flat surfaces will not get blurred by getting generated differently from two or more viewpoints. Does not use visibility masking. Each picture will get generated as new, consistency depends on IPAdapter + geometry.", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 8, "cfg": 1.5, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 80.0, "weight_exponent": 10.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "separate", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": False, "sequential_custom_camera_order": "", "sequential_factor": 0.75, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "first", "ipadapter_weight_type": "style", "ipadapter_strength": 0.800000011920929, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": False, "blur_mask": True, "blur_mask_radius": 3, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "pbr_decomposition": False, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.6000000238418579, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'sdxl_lightning_8step_lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]},+ "QWEN PRECISE": {"description": "Precise detail when camera overlap is good. Relies on context renders plus the prompt, so sparse coverage can still introduce artifacts.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit","qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": False, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "pbr_decomposition": False, "lora_units": [{'model_name': 'Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors', 'model_strength': 1.0, 'clip_strength': 0.0}]},+ "QWEN SAFE": {"description": "Safer fallback when coverage is limited. Uses the previous view (recent mode) instead of context renders to keep global look coherent, at the cost of some fine-detail persistence.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 6.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": False, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.10000000149011612, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "NONE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "pbr_decomposition": False, "lora_units": [{'model_name': 'Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors', 'model_strength': 1.0, 'clip_strength': 0.0}]},+ "QWEN ALT": {"description": "Balanced option that mixes additional context renders with sequential references to smooth out coverage while keeping detail reasonable.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": False, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "recent", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.10000000149011612, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "ADDITIONAL", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "early_priority": True, "early_priority_strength": 0.5, "pbr_decomposition": False, "lora_units": [{'model_name': 'Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors', 'model_strength': 1.0, 'clip_strength': 0.0}]},+ "QWEN VORONOI": {"description": "Voronoi projection mode with exponent 1000 for hard camera segmentation during generation, then resets to 15 for softer blending. Each surface point is dominated by its closest camera. Based on Qwen Precise.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 1000.0, "weight_exponent_generation_only": True, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": True, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "pbr_decomposition": False, "lora_units": [{'model_name': 'Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors', 'model_strength': 1.0, 'clip_strength': 0.0}]},+ "QWEN PRECISE (NUNCHAKU)": {"description": "Precise detail using Nunchaku. Meant to be used with the Nunchaku model which has the 4-step Lightning LoRA included. Requires Nunchaku nodes.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": False, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "pbr_decomposition": False, "lora_units": []},+ "QWEN SAFE (NUNCHAKU)": {"description": "Safer fallback using Nunchaku. Meant to be used with the Nunchaku model which has the 4-step Lightning LoRA included. Requires Nunchaku nodes.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 6.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": False, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.10000000149011612, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "NONE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "pbr_decomposition": False, "lora_units": []},+ "QWEN ALT (NUNCHAKU)": {"description": "Balanced option using Nunchaku. Meant to be used with the Nunchaku model which has the 4-step Lightning LoRA included. Requires Nunchaku nodes.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": False, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "recent", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.10000000149011612, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "ADDITIONAL", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "early_priority": True, "early_priority_strength": 0.5, "pbr_decomposition": False, "lora_units": []},+ "QWEN VORONOI (NUNCHAKU)": {"description": "Voronoi projection mode using Nunchaku. Exponent 1000 for hard camera segmentation during generation, then resets to 15 for softer blending. Requires Nunchaku nodes.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 1000.0, "weight_exponent_generation_only": True, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": True, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "pbr_decomposition": False, "lora_units": []},+ "KLEIN PRECISE": {"description": "FLUX.2 Klein with precise detail when camera overlap is good. Uses depth reference images with CFGGuider (cfg=1). Best for well-covered geometry.", "control_after_generate": "fixed", "model_architecture": "flux2_klein", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": False, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour.", "qwen_custom_prompt_seq_none": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour, adopting the visual style from image 2.", "qwen_custom_prompt_seq_replace": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour. In image 2, replace all solid magenta areas with content that continues the surrounding style. Replace the background with solid gray.", "qwen_custom_prompt_seq_additional": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour. In image 2, replace all solid magenta areas with content that continues the surrounding style. Replace the background with solid gray. Image 3 represents the overall style.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "pbr_decomposition": False, "lora_units": []},+ "KLEIN SAFE": {"description": "FLUX.2 Klein safer fallback when coverage is limited. Uses the previous view instead of context renders with depth references and CFGGuider (cfg=1).", "control_after_generate": "fixed", "model_architecture": "flux2_klein", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 6.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": False, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.10000000149011612, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "NONE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour.", "qwen_custom_prompt_seq_none": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour, adopting the visual style from image 2.", "qwen_custom_prompt_seq_replace": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour. In image 2, replace all solid magenta areas with content that continues the surrounding style. Replace the background with solid gray.", "qwen_custom_prompt_seq_additional": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour. In image 2, replace all solid magenta areas with content that continues the surrounding style. Replace the background with solid gray. Image 3 represents the overall style.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "pbr_decomposition": False, "lora_units": []},+ "KLEIN ALT": {"description": "FLUX.2 Klein balanced option that mixes additional context renders with sequential references. Uses depth references and CFGGuider (cfg=1).", "control_after_generate": "fixed", "model_architecture": "flux2_klein", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": False, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "recent", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.10000000149011612, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "ADDITIONAL", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour.", "qwen_custom_prompt_seq_none": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour, adopting the visual style from image 2.", "qwen_custom_prompt_seq_replace": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour. In image 2, replace all solid magenta areas with content that continues the surrounding style. Replace the background with solid gray.", "qwen_custom_prompt_seq_additional": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour. In image 2, replace all solid magenta areas with content that continues the surrounding style. Replace the background with solid gray. Image 3 represents the overall style.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "early_priority": True, "early_priority_strength": 0.5, "pbr_decomposition": False, "lora_units": []},+ "KLEIN VORONOI": {"description": "FLUX.2 Klein voronoi projection mode. Exponent 1000 for hard camera segmentation during generation, then resets to 15 for softer blending. Uses depth references and CFGGuider (cfg=1).", "control_after_generate": "fixed", "model_architecture": "flux2_klein", "qwen_generation_method": "generate", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 1000.0, "weight_exponent_generation_only": True, "weight_exponent_after_generation": 15.0, "qwen_voronoi_mode": True, "clip_skip": 1, "auto_rescale": True, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.8, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour.", "qwen_custom_prompt_seq_none": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour, adopting the visual style from image 2.", "qwen_custom_prompt_seq_replace": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour. In image 2, replace all solid magenta areas with content that continues the surrounding style. Replace the background with solid gray.", "qwen_custom_prompt_seq_additional": "Reskin this into {main_prompt}{camera_suffix}, while preserve identity keep likeness replica contour. In image 2, replace all solid magenta areas with content that continues the surrounding style. Replace the background with solid gray. Image 3 represents the overall style.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "pbr_decomposition": False, "lora_units": []},+ "LOCAL REFINE": {"description": "Uses the SDXL local edit mode to improve detail / refine specific areas. Use new set of cameras or a single camera pointed at the area you want to refine.", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 8, "cfg": 1.5, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 80.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "reinhard", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "auto_rescale_target_mp": 1.0, "overwrite_material": True, "generation_method": "local_edit", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.800000011920929, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.10000000149011612, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "original_render", "sequential_desaturate_factor": 0.10000000149011612, "sequential_contrast_factor": 0.10000000149011612, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_voronoi_mode": True, "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 5.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "refine_angle_ramp_active": True, "refine_angle_ramp_pos_0": 0.0, "refine_angle_ramp_pos_1": 0.05, "visibility_vignette": True, "visibility_vignette_width": 0.3, "visibility_vignette_softness": 1.0, "visibility_vignette_blur": False, "refine_feather_ramp_pos_0": 0.0, "refine_feather_ramp_pos_1": 0.6, "refine_edge_feather_projection": True, "refine_edge_feather_width": 30, "refine_edge_feather_softness": 1.0, "pbr_decomposition": False, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.5, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'sdxl_lightning_8step_lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]},+ "LOCAL EDIT (QWEN)": {"description": "Uses Qwen to make targeted edits to specific areas. Point cameras at what you want to change — you can alter colors, style, rewrite text, add details, and more. Untouched areas are preserved.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "qwen_generation_method": "local_edit", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "reinhard", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "auto_rescale_target_mp": 1.0, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 1.0, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": False, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_voronoi_mode": False, "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": False, "refine_angle_ramp_active": False, "refine_angle_ramp_pos_0": 0.0, "refine_angle_ramp_pos_1": 0.05, "visibility_vignette": True, "visibility_vignette_width": 0.1, "visibility_vignette_softness": 1.0, "visibility_vignette_blur": False, "refine_feather_ramp_pos_0": 0.0, "refine_feather_ramp_pos_1": 0.6, "refine_edge_feather_projection": True, "refine_edge_feather_width": 15, "refine_edge_feather_softness": 1.0, "pbr_decomposition": False, "lora_units": [{'model_name': 'Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors', 'model_strength': 1.0, 'clip_strength': 0.0}]},+ "REFINE (QWEN)": {"description": "Uses Qwen to restyle or globally modify the entire existing texture. Applies changes uniformly across all camera views — ideal for changing the overall color scheme, art style, or surface appearance.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "qwen_generation_method": "refine", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "reinhard", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "auto_rescale_target_mp": 1.0, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 1.0, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": False, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_voronoi_mode": False, "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": False, "pbr_decomposition": False, "lora_units": [{'model_name': 'Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors', 'model_strength': 1.0, 'clip_strength': 0.0}]},+ "DEFAULT (MESH + TEXTURE)": {"description": "Uses TRELLIS.2 and SDXL to generate a textured mesh. Optimized for general object generation. May not work ideally for the cases which have their specialized presets.", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 8, "cfg": 1.5, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 75.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 85.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "reinhard", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "auto_rescale_target_mp": 1.0, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.800000011920929, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.10000000149011612, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "trellis2_input", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 3, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_voronoi_mode": False, "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "NONE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 5.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "qwen_generation_method": "generate", "qwen_refine_use_prev_ref": False, "qwen_refine_use_depth": False, "qwen_timestep_zero_ref": False, "refine_angle_ramp_active": True, "refine_angle_ramp_pos_0": 0.0, "refine_angle_ramp_pos_1": 0.05000000074505806, "visibility_vignette": True, "visibility_vignette_width": 0.15000000596046448, "visibility_vignette_softness": 1.0, "visibility_vignette_blur": False, "refine_feather_ramp_pos_0": 0.0, "refine_feather_ramp_pos_1": 0.6000000238418579, "refine_edge_feather_projection": True, "refine_edge_feather_width": 30, "refine_edge_feather_softness": 1.0, "trellis2_texture_mode": "sdxl", "trellis2_initial_image_arch": "sdxl", "trellis2_camera_count": 10, "trellis2_placement_mode": "normal_weighted", "trellis2_auto_prompts": True, "trellis2_exclude_bottom": False, "trellis2_exclude_bottom_angle": 1.5533000230789185, "trellis2_auto_aspect": "per_camera", "trellis2_occlusion_mode": "none", "trellis2_consider_existing": False, "trellis2_delete_cameras": False, "trellis2_coverage_target": 0.949999988079071, "trellis2_max_auto_cameras": 12, "trellis2_fan_angle": 90.0, "trellis2_resolution": "1024_cascade", "trellis2_vram_mode": "disk_offload", "trellis2_attn_backend": "flash_attn", "trellis2_ss_guidance": 7.5, "trellis2_ss_steps": 12, "trellis2_shape_guidance": 7.5, "trellis2_shape_steps": 12, "trellis2_tex_guidance": 7.5, "trellis2_tex_steps": 12, "trellis2_max_tokens": 32768, "trellis2_texture_size": 4096, "trellis2_decimation": 1000000, "trellis2_remesh": True, "trellis2_post_processing_enabled": True, "trellis2_bg_removal": "auto", "trellis2_background_color": "black", "trellis2_fill_holes": True, "trellis2_import_scale": 2.0, "trellis2_clamp_elevation": False, "trellis2_max_elevation": 1.2216999530792236, "trellis2_min_elevation": -1.0471975803375244, "trellis2_auto_lighting": True, "use_ipadapter": False, "sequential_ipadapter_regenerate_wo_controlnet": False, "allow_modify_existing_textures": False, "ask_object_prompts": True, "weight_exponent_mask": False, "mask_blocky": False, "architecture_mode": "trellis2", "use_camera_prompts": True, "sg_use_custom_camera_order": False, "pbr_decomposition": False, "generation_mode": "standard", "texture_objects": "all", "use_flux_lora": True, "qwen_use_trellis2_style": False, "qwen_trellis2_style_initial_only": False, "trellis2_skip_texture": True, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.6000000238418579, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'sdxl_lightning_8step_lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]},+ "TRELLIS.2 (MESH ONLY)": {"description": "Uses TRELLIS.2 to generate a mesh without texturing. Useful when you only need geometry or plan to texture manually.", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 8, "cfg": 1.5, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 75.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 85.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "reinhard", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "auto_rescale_target_mp": 1.0, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.800000011920929, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.10000000149011612, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "trellis2_input", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 3, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_voronoi_mode": False, "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "NONE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 5.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "qwen_generation_method": "generate", "qwen_refine_use_prev_ref": False, "qwen_refine_use_depth": False, "qwen_timestep_zero_ref": False, "refine_angle_ramp_active": True, "refine_angle_ramp_pos_0": 0.0, "refine_angle_ramp_pos_1": 0.05000000074505806, "visibility_vignette": True, "visibility_vignette_width": 0.15000000596046448, "visibility_vignette_softness": 1.0, "visibility_vignette_blur": False, "refine_feather_ramp_pos_0": 0.0, "refine_feather_ramp_pos_1": 0.6000000238418579, "refine_edge_feather_projection": True, "refine_edge_feather_width": 30, "refine_edge_feather_softness": 1.0, "trellis2_texture_mode": "none", "trellis2_initial_image_arch": "sdxl", "trellis2_camera_count": 10, "trellis2_placement_mode": "normal_weighted", "trellis2_auto_prompts": True, "trellis2_exclude_bottom": False, "trellis2_exclude_bottom_angle": 1.5533000230789185, "trellis2_auto_aspect": "per_camera", "trellis2_occlusion_mode": "none", "trellis2_consider_existing": False, "trellis2_delete_cameras": False, "trellis2_coverage_target": 0.949999988079071, "trellis2_max_auto_cameras": 12, "trellis2_fan_angle": 90.0, "trellis2_resolution": "1024_cascade", "trellis2_vram_mode": "disk_offload", "trellis2_attn_backend": "flash_attn", "trellis2_ss_guidance": 7.5, "trellis2_ss_steps": 12, "trellis2_shape_guidance": 7.5, "trellis2_shape_steps": 12, "trellis2_tex_guidance": 7.5, "trellis2_tex_steps": 12, "trellis2_max_tokens": 32768, "trellis2_texture_size": 4096, "trellis2_decimation": 1000000, "trellis2_remesh": True, "trellis2_post_processing_enabled": True, "trellis2_bg_removal": "auto", "trellis2_background_color": "black", "trellis2_fill_holes": True, "trellis2_import_scale": 2.0, "trellis2_clamp_elevation": False, "trellis2_max_elevation": 1.2216999530792236, "trellis2_min_elevation": -1.0471975803375244, "trellis2_auto_lighting": True, "use_ipadapter": False, "sequential_ipadapter_regenerate_wo_controlnet": False, "allow_modify_existing_textures": False, "ask_object_prompts": True, "weight_exponent_mask": False, "mask_blocky": False, "architecture_mode": "trellis2", "use_camera_prompts": True, "sg_use_custom_camera_order": False, "pbr_decomposition": False, "generation_mode": "standard", "texture_objects": "all", "use_flux_lora": True, "qwen_use_trellis2_style": False, "qwen_trellis2_style_initial_only": False, "trellis2_skip_texture": True, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.6000000238418579, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'sdxl_lightning_8step_lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]},+ "CHARACTERS (MESH + TEXTURE)": {"description": "Uses TRELLIS.2 and SDXL to generate a textured mesh. Optimized settings for generating characters. Will not texture bottom facing faces.", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 8, "cfg": 1.5, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 75.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 85.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "reinhard", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "auto_rescale_target_mp": 1.0, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.800000011920929, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.10000000149011612, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "trellis2_input", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_voronoi_mode": False, "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "NONE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 5.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "qwen_generation_method": "generate", "qwen_refine_use_prev_ref": False, "qwen_refine_use_depth": False, "qwen_timestep_zero_ref": False, "refine_angle_ramp_active": True, "refine_angle_ramp_pos_0": 0.0, "refine_angle_ramp_pos_1": 0.05000000074505806, "visibility_vignette": True, "visibility_vignette_width": 0.15000000596046448, "visibility_vignette_softness": 1.0, "visibility_vignette_blur": False, "refine_feather_ramp_pos_0": 0.0, "refine_feather_ramp_pos_1": 0.6000000238418579, "refine_edge_feather_projection": True, "refine_edge_feather_width": 30, "refine_edge_feather_softness": 1.0, "trellis2_texture_mode": "sdxl", "trellis2_initial_image_arch": "sdxl", "trellis2_camera_count": 8, "trellis2_placement_mode": "normal_weighted", "trellis2_auto_prompts": True, "trellis2_exclude_bottom": True, "trellis2_exclude_bottom_angle": 1.5533000230789185, "trellis2_auto_aspect": "per_camera", "trellis2_occlusion_mode": "none", "trellis2_consider_existing": False, "trellis2_delete_cameras": False, "trellis2_coverage_target": 0.949999988079071, "trellis2_max_auto_cameras": 12, "trellis2_fan_angle": 90.0, "trellis2_resolution": "1024_cascade", "trellis2_vram_mode": "disk_offload", "trellis2_attn_backend": "flash_attn", "trellis2_ss_guidance": 7.5, "trellis2_ss_steps": 12, "trellis2_shape_guidance": 7.5, "trellis2_shape_steps": 12, "trellis2_tex_guidance": 7.5, "trellis2_tex_steps": 12, "trellis2_max_tokens": 32768, "trellis2_texture_size": 4096, "trellis2_decimation": 1000000, "trellis2_remesh": True, "trellis2_post_processing_enabled": True, "trellis2_bg_removal": "auto", "trellis2_background_color": "black", "trellis2_fill_holes": True, "trellis2_import_scale": 2.0, "trellis2_clamp_elevation": True, "trellis2_max_elevation": 0.8726646259971648, "trellis2_min_elevation": -0.8726646259971648, "trellis2_auto_lighting": True, "use_ipadapter": False, "sequential_ipadapter_regenerate_wo_controlnet": False, "allow_modify_existing_textures": False, "ask_object_prompts": True, "weight_exponent_mask": False, "mask_blocky": False, "architecture_mode": "trellis2", "use_camera_prompts": True, "sg_use_custom_camera_order": False, "pbr_decomposition": False, "generation_mode": "standard", "texture_objects": "all", "use_flux_lora": True, "qwen_use_trellis2_style": False, "qwen_trellis2_style_initial_only": False, "trellis2_skip_texture": True, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.6000000238418579, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'sdxl_lightning_8step_lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]},+ "ARCHITECTURE (MESH + TEXTURE)": {"description": "Uses TRELLIS.2 and SDXL to generate a textured mesh. Optimized settings for architecture, and other models with flat walls and sharp angles.", "control_after_generate": "fixed", "model_architecture": "sdxl", "steps": 8, "cfg": 1.5, "sampler": "dpmpp_2s_ancestral", "scheduler": "sgm_uniform", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 80.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 85.0, "weight_exponent": 10.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "reinhard", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "auto_rescale_target_mp": 1.0, "overwrite_material": True, "generation_method": "separate", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.800000011920929, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": False, "sequential_custom_camera_order": "", "sequential_factor": 0.75, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": True, "sequential_ipadapter_mode": "trellis2_input", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 0.800000011920929, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": False, "blur_mask": True, "blur_mask_radius": 3, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_voronoi_mode": False, "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "NONE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 5.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "qwen_generation_method": "generate", "qwen_refine_use_prev_ref": False, "qwen_refine_use_depth": False, "qwen_timestep_zero_ref": False, "refine_angle_ramp_active": True, "refine_angle_ramp_pos_0": 0.0, "refine_angle_ramp_pos_1": 0.05000000074505806, "visibility_vignette": True, "visibility_vignette_width": 0.15000000596046448, "visibility_vignette_softness": 1.0, "visibility_vignette_blur": False, "refine_feather_ramp_pos_0": 0.0, "refine_feather_ramp_pos_1": 0.6000000238418579, "refine_edge_feather_projection": True, "refine_edge_feather_width": 30, "refine_edge_feather_softness": 1.0, "trellis2_texture_mode": "sdxl", "trellis2_initial_image_arch": "sdxl", "trellis2_camera_count": 8, "trellis2_placement_mode": "normal_weighted", "trellis2_auto_prompts": True, "trellis2_exclude_bottom": True, "trellis2_exclude_bottom_angle": 1.5533000230789185, "trellis2_auto_aspect": "per_camera", "trellis2_occlusion_mode": "none", "trellis2_consider_existing": False, "trellis2_delete_cameras": False, "trellis2_coverage_target": 0.949999988079071, "trellis2_max_auto_cameras": 12, "trellis2_fan_angle": 90.0, "trellis2_resolution": "1024_cascade", "trellis2_vram_mode": "disk_offload", "trellis2_attn_backend": "flash_attn", "trellis2_ss_guidance": 7.5, "trellis2_ss_steps": 12, "trellis2_shape_guidance": 7.5, "trellis2_shape_steps": 12, "trellis2_tex_guidance": 7.5, "trellis2_tex_steps": 12, "trellis2_max_tokens": 32768, "trellis2_texture_size": 4096, "trellis2_decimation": 1000000, "trellis2_remesh": True, "trellis2_post_processing_enabled": True, "trellis2_bg_removal": "auto", "trellis2_background_color": "black", "trellis2_fill_holes": True, "trellis2_import_scale": 2.0, "trellis2_clamp_elevation": False, "trellis2_max_elevation": 1.2216999530792236, "trellis2_min_elevation": -1.0471975803375244, "trellis2_auto_lighting": True, "use_ipadapter": False, "sequential_ipadapter_regenerate_wo_controlnet": False, "allow_modify_existing_textures": False, "ask_object_prompts": True, "weight_exponent_mask": False, "mask_blocky": False, "architecture_mode": "trellis2", "use_camera_prompts": True, "sg_use_custom_camera_order": False, "pbr_decomposition": False, "generation_mode": "standard", "texture_objects": "all", "use_flux_lora": True, "qwen_use_trellis2_style": False, "qwen_trellis2_style_initial_only": False, "trellis2_skip_texture": True, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.6000000238418579, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'sdxl_lightning_8step_lora.safetensors', 'model_strength': 1.0, 'clip_strength': 1.0}]},+ "QWEN PRECISE (MESH + TEXTURE)": {"description": "Uses TRELLIS.2 and Qwen Image Edit to generate a textured mesh. Precise detail when camera overlap is good. Relies on context renders plus the prompt.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "reinhard", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "auto_rescale_target_mp": 1.0, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.800000011920929, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_voronoi_mode": False, "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "qwen_generation_method": "generate", "qwen_refine_use_prev_ref": False, "qwen_refine_use_depth": False, "qwen_timestep_zero_ref": False, "refine_angle_ramp_active": True, "refine_angle_ramp_pos_0": 0.0, "refine_angle_ramp_pos_1": 0.05000000074505806, "visibility_vignette": True, "visibility_vignette_width": 0.15000000596046448, "visibility_vignette_softness": 1.0, "visibility_vignette_blur": False, "sg_silhouette_margin": 3, "sg_silhouette_depth": 0.05000000074505806, "sg_silhouette_rays": "4", "refine_feather_ramp_pos_0": 0.0, "refine_feather_ramp_pos_1": 0.6000000238418579, "refine_edge_feather_projection": True, "refine_edge_feather_width": 30, "refine_edge_feather_softness": 1.0, "trellis2_texture_mode": "qwen_image_edit", "trellis2_initial_image_arch": "sdxl", "trellis2_camera_count": 8, "trellis2_placement_mode": "normal_weighted", "trellis2_auto_prompts": True, "trellis2_exclude_bottom": True, "trellis2_exclude_bottom_angle": 1.5533000230789185, "trellis2_auto_aspect": "per_camera", "trellis2_occlusion_mode": "none", "trellis2_consider_existing": False, "trellis2_delete_cameras": False, "trellis2_coverage_target": 0.949999988079071, "trellis2_max_auto_cameras": 12, "trellis2_fan_angle": 90.0, "trellis2_resolution": "1024_cascade", "trellis2_vram_mode": "disk_offload", "trellis2_attn_backend": "flash_attn", "trellis2_seed": 0, "trellis2_ss_guidance": 7.5, "trellis2_ss_steps": 12, "trellis2_shape_guidance": 7.5, "trellis2_shape_steps": 12, "trellis2_tex_guidance": 7.5, "trellis2_tex_steps": 12, "trellis2_max_tokens": 32768, "trellis2_texture_size": 4096, "trellis2_decimation": 1000000, "trellis2_remesh": True, "trellis2_post_processing_enabled": True, "trellis2_bg_removal": "auto", "trellis2_background_color": "black", "trellis2_fill_holes": True, "trellis2_import_scale": 2.0, "trellis2_clamp_elevation": True, "trellis2_max_elevation": 0.8726646259971648, "trellis2_min_elevation": -0.8726646259971648, "trellis2_auto_lighting": True, "use_ipadapter": False, "sequential_ipadapter_regenerate_wo_controlnet": False, "allow_modify_existing_textures": False, "ask_object_prompts": True, "weight_exponent_mask": False, "mask_blocky": False, "architecture_mode": "trellis2", "use_camera_prompts": True, "sg_use_custom_camera_order": False, "pbr_decomposition": False, "generation_mode": "standard", "texture_objects": "all", "use_flux_lora": True, "qwen_use_trellis2_style": True, "qwen_trellis2_style_initial_only": True, "trellis2_skip_texture": True, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.6000000238418579, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors', 'model_strength': 1.0, 'clip_strength': 0.0}]},+ "QWEN SAFE (MESH + TEXTURE)": {"description": "Uses TRELLIS.2 and Qwen Image Edit to generate a textured mesh. Safer fallback when coverage is limited. Uses TRELLIS.2 input image as sequential style reference for global coherence.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 6.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "reinhard", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "auto_rescale_target_mp": 1.0, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.800000011920929, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.10000000149011612, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_voronoi_mode": False, "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "NONE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "qwen_generation_method": "generate", "qwen_refine_use_prev_ref": False, "qwen_refine_use_depth": False, "qwen_timestep_zero_ref": False, "refine_angle_ramp_active": True, "refine_angle_ramp_pos_0": 0.0, "refine_angle_ramp_pos_1": 0.05000000074505806, "visibility_vignette": True, "visibility_vignette_width": 0.15000000596046448, "visibility_vignette_softness": 1.0, "visibility_vignette_blur": False, "sg_silhouette_margin": 3, "sg_silhouette_depth": 0.05000000074505806, "sg_silhouette_rays": "4", "refine_feather_ramp_pos_0": 0.0, "refine_feather_ramp_pos_1": 0.6000000238418579, "refine_edge_feather_projection": True, "refine_edge_feather_width": 30, "refine_edge_feather_softness": 1.0, "trellis2_texture_mode": "qwen_image_edit", "trellis2_initial_image_arch": "sdxl", "trellis2_camera_count": 8, "trellis2_placement_mode": "normal_weighted", "trellis2_auto_prompts": True, "trellis2_exclude_bottom": True, "trellis2_exclude_bottom_angle": 1.5533000230789185, "trellis2_auto_aspect": "per_camera", "trellis2_occlusion_mode": "none", "trellis2_consider_existing": False, "trellis2_delete_cameras": False, "trellis2_coverage_target": 0.949999988079071, "trellis2_max_auto_cameras": 12, "trellis2_fan_angle": 90.0, "trellis2_resolution": "1024_cascade", "trellis2_vram_mode": "disk_offload", "trellis2_attn_backend": "flash_attn", "trellis2_seed": 0, "trellis2_ss_guidance": 7.5, "trellis2_ss_steps": 12, "trellis2_shape_guidance": 7.5, "trellis2_shape_steps": 12, "trellis2_tex_guidance": 7.5, "trellis2_tex_steps": 12, "trellis2_max_tokens": 32768, "trellis2_texture_size": 4096, "trellis2_decimation": 1000000, "trellis2_remesh": True, "trellis2_post_processing_enabled": True, "trellis2_bg_removal": "auto", "trellis2_background_color": "black", "trellis2_fill_holes": True, "trellis2_import_scale": 2.0, "trellis2_clamp_elevation": True, "trellis2_max_elevation": 0.8726646259971648, "trellis2_min_elevation": -0.8726646259971648, "trellis2_auto_lighting": True, "use_ipadapter": False, "sequential_ipadapter_regenerate_wo_controlnet": False, "allow_modify_existing_textures": False, "ask_object_prompts": True, "weight_exponent_mask": False, "mask_blocky": False, "architecture_mode": "trellis2", "use_camera_prompts": True, "sg_use_custom_camera_order": False, "pbr_decomposition": False, "generation_mode": "standard", "texture_objects": "all", "use_flux_lora": True, "qwen_use_trellis2_style": True, "qwen_trellis2_style_initial_only": False, "trellis2_skip_texture": True, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.6000000238418579, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors', 'model_strength': 1.0, 'clip_strength': 0.0}]},+ "QWEN ALT (MESH + TEXTURE)": {"description": "Uses TRELLIS.2 and Qwen Image Edit to generate a textured mesh. Uses ADDITIONAL context renders with TRELLIS.2 style transfer for consistency. Good general-purpose Qwen pipeline.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 3.0, "weight_exponent_generation_only": False, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "reinhard", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "auto_rescale_target_mp": 1.0, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.800000011920929, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.10000000149011612, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": True, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_voronoi_mode": False, "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "ADDITIONAL", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "qwen_generation_method": "generate", "qwen_refine_use_prev_ref": False, "qwen_refine_use_depth": False, "qwen_timestep_zero_ref": False, "refine_angle_ramp_active": True, "refine_angle_ramp_pos_0": 0.0, "refine_angle_ramp_pos_1": 0.05000000074505806, "visibility_vignette": True, "visibility_vignette_width": 0.15000000596046448, "visibility_vignette_softness": 1.0, "visibility_vignette_blur": False, "sg_silhouette_margin": 3, "sg_silhouette_depth": 0.05000000074505806, "sg_silhouette_rays": "4", "refine_feather_ramp_pos_0": 0.0, "refine_feather_ramp_pos_1": 0.6000000238418579, "refine_edge_feather_projection": True, "refine_edge_feather_width": 30, "refine_edge_feather_softness": 1.0, "trellis2_texture_mode": "qwen_image_edit", "trellis2_initial_image_arch": "sdxl", "trellis2_camera_count": 8, "trellis2_placement_mode": "normal_weighted", "trellis2_auto_prompts": True, "trellis2_exclude_bottom": True, "trellis2_exclude_bottom_angle": 1.5533000230789185, "trellis2_auto_aspect": "per_camera", "trellis2_occlusion_mode": "none", "trellis2_consider_existing": False, "trellis2_delete_cameras": False, "trellis2_coverage_target": 0.949999988079071, "trellis2_max_auto_cameras": 12, "trellis2_fan_angle": 90.0, "trellis2_resolution": "1024_cascade", "trellis2_vram_mode": "disk_offload", "trellis2_attn_backend": "flash_attn", "trellis2_seed": 0, "trellis2_ss_guidance": 7.5, "trellis2_ss_steps": 12, "trellis2_shape_guidance": 7.5, "trellis2_shape_steps": 12, "trellis2_tex_guidance": 7.5, "trellis2_tex_steps": 12, "trellis2_max_tokens": 32768, "trellis2_texture_size": 4096, "trellis2_decimation": 1000000, "trellis2_remesh": True, "trellis2_post_processing_enabled": True, "trellis2_bg_removal": "auto", "trellis2_background_color": "black", "trellis2_fill_holes": True, "trellis2_import_scale": 2.0, "trellis2_clamp_elevation": True, "trellis2_max_elevation": 0.8726646259971648, "trellis2_min_elevation": -0.8726646259971648, "trellis2_auto_lighting": True, "use_ipadapter": False, "sequential_ipadapter_regenerate_wo_controlnet": False, "allow_modify_existing_textures": False, "ask_object_prompts": True, "weight_exponent_mask": False, "mask_blocky": False, "architecture_mode": "trellis2", "use_camera_prompts": True, "sg_use_custom_camera_order": False, "pbr_decomposition": False, "generation_mode": "standard", "texture_objects": "all", "use_flux_lora": True, "qwen_use_trellis2_style": True, "qwen_trellis2_style_initial_only": False, "trellis2_skip_texture": True, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.6000000238418579, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors', 'model_strength': 1.0, 'clip_strength': 0.0}]},+ "QWEN VORONOI (MESH + TEXTURE)": {"description": "Uses TRELLIS.2 and Qwen Image Edit to generate a textured mesh. Voronoi projection with exponent 1000 for hard camera segmentation during generation, then resets to 15 for softer blending.", "control_after_generate": "fixed", "model_architecture": "qwen_image_edit", "steps": 4, "cfg": 1.0, "sampler": "euler", "scheduler": "simple", "fallback_color": mathutils.Color((0.0, 0.0, 0.0)), "discard_factor": 70.0, "discard_factor_generation_only": True, "discard_factor_after_generation": 90.0, "weight_exponent": 1000.0, "weight_exponent_generation_only": True, "weight_exponent_after_generation": 15.0, "view_blend_use_color_match": False, "view_blend_color_match_method": "reinhard", "view_blend_color_match_strength": 1.0, "clip_skip": 1, "auto_rescale": True, "auto_rescale_target_mp": 1.0, "overwrite_material": True, "generation_method": "sequential", "refine_images": False, "refine_steps": 8, "refine_sampler": "dpmpp_2s_ancestral", "refine_scheduler": "sgm_uniform", "denoise": 0.800000011920929, "refine_cfg": 1.5, "refine_prompt": "", "refine_upscale_method": "lanczos", "sequential_smooth": True, "sequential_custom_camera_order": "", "sequential_factor": 0.699999988079071, "sequential_factor_smooth": 0.15000000596046448, "sequential_factor_smooth_2": 1.0, "sequential_ipadapter": False, "sequential_ipadapter_mode": "first", "sequential_desaturate_factor": 0.0, "sequential_contrast_factor": 0.0, "sequential_ipadapter_regenerate": False, "ipadapter_weight_type": "style", "ipadapter_strength": 1.0, "ipadapter_start": 0.0, "ipadapter_end": 1.0, "early_priority": False, "early_priority_strength": 0.5, "differential_diffusion": True, "differential_noise": True, "blur_mask": True, "blur_mask_radius": 1, "blur_mask_sigma": 1.0, "grow_mask_by": 3, "canny_threshold_low": 0, "canny_threshold_high": 80, "qwen_guidance_map_type": "depth", "qwen_voronoi_mode": True, "qwen_use_external_style_image": False, "qwen_external_style_image": "", "qwen_context_render_mode": "REPLACE_STYLE", "qwen_external_style_initial_only": False, "qwen_use_custom_prompts": False, "qwen_custom_prompt_initial": "Change the format of image 1 to '{main_prompt}'", "qwen_custom_prompt_seq_none": "Change and transfer the format of '{main_prompt}' in image 1 to the style from image 2", "qwen_custom_prompt_seq_replace": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas.", "qwen_custom_prompt_seq_additional": "Change and transfer the format of image 1 to '{main_prompt}'. Replace all solid magenta areas in image 2. Replace the background with solid gray. The style from image 2 should smoothly continue into the previously magenta areas. Image 3 represents the overall style of the object.", "qwen_guidance_fallback_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_guidance_background_color": mathutils.Color((1.0, 0.0, 1.0)), "qwen_context_cleanup": False, "qwen_context_cleanup_hue_tolerance": 0.0, "qwen_context_cleanup_value_adjust": 0.0, "qwen_context_fallback_dilation": 1, "qwen_rescale_alignment": True, "qwen_generation_method": "generate", "qwen_refine_use_prev_ref": False, "qwen_refine_use_depth": False, "qwen_timestep_zero_ref": False, "refine_angle_ramp_active": True, "refine_angle_ramp_pos_0": 0.0, "refine_angle_ramp_pos_1": 0.05000000074505806, "visibility_vignette": True, "visibility_vignette_width": 0.15000000596046448, "visibility_vignette_softness": 1.0, "visibility_vignette_blur": False, "sg_silhouette_margin": 3, "sg_silhouette_depth": 0.05000000074505806, "sg_silhouette_rays": "4", "refine_feather_ramp_pos_0": 0.0, "refine_feather_ramp_pos_1": 0.6000000238418579, "refine_edge_feather_projection": True, "refine_edge_feather_width": 30, "refine_edge_feather_softness": 1.0, "trellis2_texture_mode": "qwen_image_edit", "trellis2_initial_image_arch": "sdxl", "trellis2_camera_count": 8, "trellis2_placement_mode": "normal_weighted", "trellis2_auto_prompts": True, "trellis2_exclude_bottom": True, "trellis2_exclude_bottom_angle": 1.5533000230789185, "trellis2_auto_aspect": "per_camera", "trellis2_occlusion_mode": "none", "trellis2_consider_existing": False, "trellis2_delete_cameras": False, "trellis2_coverage_target": 0.949999988079071, "trellis2_max_auto_cameras": 12, "trellis2_fan_angle": 90.0, "trellis2_resolution": "1024_cascade", "trellis2_vram_mode": "disk_offload", "trellis2_attn_backend": "flash_attn", "trellis2_seed": 0, "trellis2_ss_guidance": 7.5, "trellis2_ss_steps": 12, "trellis2_shape_guidance": 7.5, "trellis2_shape_steps": 12, "trellis2_tex_guidance": 7.5, "trellis2_tex_steps": 12, "trellis2_max_tokens": 32768, "trellis2_texture_size": 4096, "trellis2_decimation": 1000000, "trellis2_remesh": True, "trellis2_post_processing_enabled": True, "trellis2_bg_removal": "auto", "trellis2_background_color": "black", "trellis2_fill_holes": True, "trellis2_import_scale": 2.0, "trellis2_clamp_elevation": True, "trellis2_max_elevation": 0.8726646259971648, "trellis2_min_elevation": -0.8726646259971648, "trellis2_auto_lighting": True, "use_ipadapter": False, "sequential_ipadapter_regenerate_wo_controlnet": False, "allow_modify_existing_textures": False, "ask_object_prompts": True, "weight_exponent_mask": False, "mask_blocky": False, "architecture_mode": "trellis2", "use_camera_prompts": True, "sg_use_custom_camera_order": False, "pbr_decomposition": False, "generation_mode": "standard", "texture_objects": "all", "use_flux_lora": True, "qwen_use_trellis2_style": True, "qwen_trellis2_style_initial_only": False, "trellis2_skip_texture": True, "controlnet_units": [{'unit_type': 'depth', 'model_name': 'controlnet_depth_sdxl.safetensors', 'strength': 0.6000000238418579, 'start_percent': 0.0, 'end_percent': 1.0, 'is_union': False, 'use_union_type': True}], "lora_units": [{'model_name': 'Qwen-Image-Edit-2509-Lightning-4steps-V1.0-bf16.safetensors', 'model_strength': 1.0, 'clip_strength': 0.0}]},}# Global list of all generation parameter names to check for a preset.@@ -167,7 +169,7 @@ GEN_PARAMETERS = ["trellis2_decimation","trellis2_remesh","trellis2_post_processing_enabled",- "trellis2_low_vram",+"trellis2_bg_removal","trellis2_background_color","trellis2_fill_holes",@@ -187,13 +189,13 @@ GEN_PARAMETERS = ["architecture_mode","use_camera_prompts","sg_use_custom_camera_order",- "apply_bsdf","generation_mode","texture_objects","use_flux_lora","qwen_use_trellis2_style","qwen_trellis2_style_initial_only","trellis2_skip_texture",+ "pbr_decomposition",]def get_preset_items(self, context):@@ -424,7 +426,10 @@ class StableGenPanel(bpy.types.Panel):pbr_active = getattr(comfy_tex_op, '_pbr_active', False)if pbr_active:- # PBR decomposition sub-phase+ # PBR decomposition sub-phase — 3 bars:+ # Bar 1 (top): overall pipeline progress+ # Bar 2: current PBR model (camera progress)+ # Bar 3: PBR step N/Mpbr_step = getattr(comfy_tex_op, '_pbr_step', 0)pbr_total = max(getattr(comfy_tex_op, '_pbr_total_steps', 1), 1)pbr_cam = getattr(comfy_tex_op, '_pbr_cam', 0)@@ -432,11 +437,25 @@ class StableGenPanel(bpy.types.Panel):cam_frac = pbr_cam / pbr_cam_totalstage_text = getattr(comfy_tex_op, '_stage', 'PBR Decomposition')- # Bar 1 — Current model (camera progress)+ # Overall: texturing is done, PBR owns the upper slice of Phase 3+ phase_start = getattr(scene, 'trellis2_pipeline_phase_start_pct', 65.0)+ phase_weight = 100.0 - phase_start+ tex_portion = phase_weight * 0.6 # first 60% of phase = texturing+ pbr_portion = phase_weight * 0.4 # last 40% = PBR+ pbr_frac = max(0.0, min(((pbr_step - 1) + cam_frac) / pbr_total, 1.0))+ overall_pct = phase_start + tex_portion + pbr_portion * pbr_frac+ overall_pct = max(0.0, min(overall_pct, 100.0))+ total_phases = getattr(scene, 'trellis2_pipeline_total_phases', 3)++ # Bar 1 — Overall pipeline+ progress_col.progress(+ text=f"Phase {total_phases}/{total_phases}: PBR Decomposition ({overall_pct:.0f}%)",+ factor=max(0.0, min(overall_pct / 100.0, 1.0))+ )+ # Bar 2 — Current model (camera progress)progress_col.progress(text=stage_text, factor=max(0.0, min(cam_frac, 1.0)))- # Bar 2 — PBR step N/M- pbr_factor = max(0.0, min(((pbr_step - 1) + cam_frac) / pbr_total, 1.0))- progress_col.progress(text=f"PBR: Step {pbr_step}/{pbr_total}", factor=pbr_factor)+ # Bar 3 — PBR step N/M+ progress_col.progress(text=f"PBR: Step {pbr_step}/{pbr_total}", factor=pbr_frac)else:# Normal texturing progress# Compute overall from scene pipeline props + ComfyUI progress@@ -710,8 +729,10 @@ class StableGenPanel(bpy.types.Panel):gif_col.prop(wm, "sg_queue_gif_hdri_strength")gif_col.prop(wm, "sg_queue_gif_hdri_rotation")gif_col.prop(wm, "sg_queue_gif_env_mode")- gif_col.prop(wm, "sg_queue_gif_shadow_plane")gif_col.prop(wm, "sg_queue_gif_denoiser")+ if bpy.app.version >= (5, 1, 0) and getattr(wm, 'sg_queue_gif_engine', 'CYCLES') == 'CYCLES':+ gif_col.prop(wm, "sg_queue_gif_use_gpu")+ gif_col.prop(wm, "sg_queue_gif_also_no_pbr")# --- Main Parameters section ---if not hasattr(scene, 'show_generation_params'):@@ -729,7 +750,15 @@ class StableGenPanel(bpy.types.Panel):# Split for promptsplit = params_container.split(factor=0.25)split.label(text="Prompt:")- split.prop(scene, "comfyui_prompt", text="")+ prompt_row = split.row(align=True)+ prompt_row.prop(scene, "comfyui_prompt", text="")+ prompt_row.prop(scene, "use_separate_texture_prompt", text="",+ icon='BRUSH_DATA',+ icon_only=True)+ if scene.use_separate_texture_prompt:+ split = params_container.split(factor=0.25)+ split.label(text="Texture Prompt:")+ split.prop(scene, "texture_prompt", text="")# Architecture selector (architecture_mode — includes TRELLIS.2)split = params_container.split(factor=0.5)@@ -917,10 +946,6 @@ class StableGenPanel(bpy.types.Panel):split = content_box.split(factor=0.5)split.label(text="BG Removal:")split.prop(scene, "trellis2_bg_removal", text="")- if scene.trellis2_bg_removal == 'auto':- row = content_box.row()- row.prop(scene, "trellis2_low_vram", text="Low VRAM Background Removal", toggle=True, icon="GHOST_ENABLED")-content_box.separator()# Artifact Filtering (nested collapsible)@@ -932,6 +957,8 @@ class StableGenPanel(bpy.types.Panel):row.prop(scene, "trellis2_artifact_spike_abs_max", text="Max Spike Count")row = af_box.row()row.prop(scene, "trellis2_artifact_max_retries", text="Max Retries")+ row = af_box.row()+ row.prop(scene, "trellis2_artifact_fail_on_exhausted", text="Fail if Unresolved", toggle=True, icon="CANCEL")# --- TRELLIS.2: Native Texture Settings ---if is_trellis2 and trellis2_tex_mode == 'native':@@ -1185,32 +1212,12 @@ class StableGenPanel(bpy.types.Panel):else:content_box = Noneif content_box:- split = content_box.split(factor=0.5)- split.label(text="Fallback Color:")- split.prop(scene, "fallback_color", text="")--- row = content_box.row()- row.prop(scene, "apply_bsdf", text="Apply BSDF", toggle=True, icon="SHADING_RENDERED")-- row = content_box.row()- row.prop(scene, "auto_rescale", text="Auto Rescale Resolution", toggle=True, icon="ARROW_LEFTRIGHT")- if scene.auto_rescale:- sub_box_rescale = content_box.box()- row = sub_box_rescale.row()- row.prop(scene, "auto_rescale_target_mp", text="Target Megapixels")- if scene.model_architecture.startswith('qwen'):- row = sub_box_rescale.row()- row.prop(scene, "qwen_rescale_alignment", text="Qwen VL-Aligned Rescale (112px)", toggle=True, icon="SNAP_INCREMENT")- row = content_box.row()- row.prop(scene, "overwrite_material", text="Overwrite Material", toggle=True, icon="FILE_REFRESH")-# ── PBR Decomposition ──- content_box.separator()row = content_box.row()row.prop(scene, "pbr_decomposition", text="PBR Decomposition", toggle=True, icon="NODE_MATERIAL")if scene.pbr_decomposition:sub = content_box.box()+ sub.label(text="PBR Decomposition", icon="NODE_MATERIAL")# ── Map toggles ──sub.label(text="Maps to Extract:", icon="IMAGE_DATA")grid = sub.grid_flow(row_major=True, columns=3, even_columns=True, align=True)@@ -1223,8 +1230,6 @@ class StableGenPanel(bpy.types.Panel):grid.prop(scene, "pbr_map_emission", toggle=True, icon="LIGHT_POINT")# ── Per-map adjustment controls ──if scene.pbr_map_normal:- adj_row = sub.row()- adj_row.prop(scene, "pbr_normal_mode", text="Normal Mode")adj_row = sub.row()adj_row.prop(scene, "pbr_normal_strength", text="Normal Strength", slider=True)if scene.pbr_map_height:@@ -1247,9 +1252,6 @@ class StableGenPanel(bpy.types.Panel):hsv_row.prop(scene, "pbr_emission_value_min", text="Val Min", slider=True)hsv_row = sub.row()hsv_row.prop(scene, "pbr_emission_bloom", text="Bloom Radius", slider=True)- elif scene.pbr_emission_method == 'vlm_seg':- kw_row = sub.row()- kw_row.prop(scene, "pbr_emission_keywords", text="Keywords")# ── Albedo source selector (only when albedo enabled) ──if scene.pbr_map_albedo:sub.separator()@@ -1258,6 +1260,10 @@ class StableGenPanel(bpy.types.Panel):if scene.pbr_albedo_source == 'delight':adj_row = sub.row()adj_row.prop(scene, "pbr_delight_strength", text="Delight Strength", slider=True)+ adj_row = sub.row(align=True)+ adj_row.prop(scene, "pbr_albedo_auto_saturation", text="Correct Albedo Saturation", toggle=True, icon="BRUSHES_ALL")+ if scene.pbr_albedo_auto_saturation:+ adj_row.prop(scene, "pbr_albedo_per_camera_saturation", text="Per Camera", toggle=True)sub.separator()# ── Quality settings ──row = sub.row(align=True)@@ -1265,7 +1271,7 @@ class StableGenPanel(bpy.types.Panel):row.prop(scene, "pbr_tiling", text="")if scene.pbr_tiling != 'off':row = sub.row(align=True)- row.prop(scene, "pbr_tile_grid", text="Tile Grid (N×N)")+ row.prop(scene, "pbr_tile_grid", text="Tile Grid (N\u00d7N)")row.prop(scene, "pbr_tile_superres", text="Super Res", toggle=True, icon="IMAGE_PLANE")if scene.pbr_tiling == 'custom':row = sub.row(align=True)@@ -1286,6 +1292,23 @@ class StableGenPanel(bpy.types.Panel):row = sub.row()row.prop(scene, "pbr_auto_lighting", text="Studio Lighting", toggle=True, icon="LIGHT_AREA")+ content_box.separator()+ split = content_box.split(factor=0.5)+ split.label(text="Fallback Color:")+ split.prop(scene, "fallback_color", text="")++ row = content_box.row()+ row.prop(scene, "auto_rescale", text="Auto Rescale Resolution", toggle=True, icon="ARROW_LEFTRIGHT")+ if scene.auto_rescale:+ sub_box_rescale = content_box.box()+ row = sub_box_rescale.row()+ row.prop(scene, "auto_rescale_target_mp", text="Target Megapixels")+ if scene.model_architecture.startswith('qwen'):+ row = sub_box_rescale.row()+ row.prop(scene, "qwen_rescale_alignment", text="Qwen VL-Aligned Rescale (112px)", toggle=True, icon="SNAP_INCREMENT")+ row = content_box.row()+ row.prop(scene, "overwrite_material", text="Overwrite Material", toggle=True, icon="FILE_REFRESH")+# --- Image Guidance (IPAdapter & ControlNet) ---if _show_diffusion_sections:if scene.model_architecture in ['sdxl', 'flux1']:@@ -1948,6 +1971,13 @@ class StableGenPanel(bpy.types.Panel):row = debug_box.row()op2 = row.operator("stablegen.debug_feather_preview", text="Feather Preview", icon="MOD_SMOOTH")+ # --- Narrow panel warning ---+ if width_mode == 'narrow':+ warn_box = layout.box()+ warn_row = warn_box.row()+ warn_row.alert = True+ warn_row.label(text="Panel too narrow — widen for full layout", icon="ERROR")+layout.separator()@@ -1959,6 +1989,10 @@ class ResetQwenPrompt(bpy.types.Operator):prompt_type: bpy.props.StringProperty()+ @classmethod+ def poll(cls, context):+ return not sg_modal_active(context)+def execute(self, context):from .workflows import WorkflowManager# We need an operator instance to call the helper, a bit of a workaround@@ -1992,6 +2026,10 @@ class ApplyPreset(bpy.types.Operator):bl_label = "Apply Preset"bl_description = "Set multiple parameters based on selected preset for easier configuration"+ @classmethod+ def poll(cls, context):+ return not sg_modal_active(context)+def execute(self, context):preset = context.scene.stablegen_presetif preset in PRESETS:@@ -2072,7 +2110,11 @@ class SavePreset(bpy.types.Operator):"""Save the current parameter values as a custom preset"""bl_idname = "stablegen.save_preset"bl_label = "Save Custom Preset"-++ @classmethod+ def poll(cls, context):+ return not sg_modal_active(context)+preset_name: bpy.props.StringProperty(name="Preset Name",default="MyPreset"@@ -2179,6 +2221,10 @@ class DeletePreset(bpy.types.Operator):bl_idname = "stablegen.delete_preset"bl_label = "Delete Preset"+ @classmethod+ def poll(cls, context):+ return not sg_modal_active(context)+def execute(self, context):preset = context.scene.stablegen_presetif preset in PRESETS:@@ -2208,6 +2254,10 @@ _queue_phase = 'idle' # 'idle' | 'switching' | 'running' | 'settling' | '_queue_settle_deadline = 0.0 # monotonic time after which 'settling' may finish_queue_force_reload = False # Force re-open of the .blend on next idle tick (for retries)_queue_texturing_seen = False # Set True once generation_status='running' is observed during settling+_queue_refresh_deadline = 0.0 # monotonic deadline while waiting for model-list refresh+_queue_original_pbr = False # stored pbr_decomposition before no-PBR reproject+_queue_original_gen_method = 'sequential' # stored generation_method before reproject+_queue_original_overwrite = True # stored overwrite_material before reproject_QUEUE_MAX_RETRIES = 3 # how many times a failed item is retried@@ -2295,7 +2345,7 @@ class SceneQueueAdd(bpy.types.Operator):@classmethoddef poll(cls, context):- return not _queue_processing+ return not _queue_processing and not sg_modal_active(context)def invoke(self, context, event):scene = context.scene@@ -2339,6 +2389,10 @@ class SceneQueueAdd(bpy.types.Operator):copy_path = os.path.join(jobs_dir, copy_name)# Save a copy — current file stays active & untouched+ # Snapshot the model_name into the plain-string backup property+ # BEFORE saving so the copy carries the correct value even if+ # the dynamic Enum resolves wrong on re-open (stale item cache).+ scene.sg_model_name_backup = scene.model_nametry:bpy.ops.wm.save_as_mainfile(filepath=copy_path, copy=True)except Exception as e:@@ -2367,7 +2421,7 @@ class SceneQueueRemove(bpy.types.Operator):@classmethoddef poll(cls, context):wm = context.window_manager- return not _queue_processing and len(wm.sg_scene_queue) > 0+ return not _queue_processing and len(wm.sg_scene_queue) > 0 and not sg_modal_active(context)def invoke(self, context, event):wm = context.window_manager@@ -2411,7 +2465,7 @@ class SceneQueueClear(bpy.types.Operator):@classmethoddef poll(cls, context):wm = context.window_manager- return not _queue_processing and len(wm.sg_scene_queue) > 0+ return not _queue_processing and len(wm.sg_scene_queue) > 0 and not sg_modal_active(context)def invoke(self, context, event):wm = context.window_manager@@ -2446,7 +2500,7 @@ class SceneQueueMoveUp(bpy.types.Operator):@classmethoddef poll(cls, context):wm = context.window_manager- return not _queue_processing and wm.sg_scene_queue_index > 0+ return not _queue_processing and wm.sg_scene_queue_index > 0 and not sg_modal_active(context)def execute(self, context):wm = context.window_manager@@ -2465,7 +2519,8 @@ class SceneQueueMoveDown(bpy.types.Operator):def poll(cls, context):wm = context.window_managerreturn (not _queue_processing- and wm.sg_scene_queue_index < len(wm.sg_scene_queue) - 1)+ and wm.sg_scene_queue_index < len(wm.sg_scene_queue) - 1+ and not sg_modal_active(context))def execute(self, context):wm = context.window_manager@@ -2513,7 +2568,7 @@ class SceneQueueProcess(bpy.types.Operator):wm = context.window_managerif _queue_processing:return True # Allow clicking to cancel- return len(wm.sg_scene_queue) > 0+ return len(wm.sg_scene_queue) > 0 and not sg_modal_active(context)def execute(self, context):global _queue_processing, _queue_timer, _queue_current_idx, _queue_phase@@ -2563,6 +2618,7 @@ def _queue_tick():import time as _timeglobal _queue_processing, _queue_current_idx, _queue_phase, _queue_timerglobal _queue_settle_deadline, _queue_force_reload, _queue_texturing_seen+ global _queue_refresh_deadlineif not _queue_processing:_queue_phase = 'idle'@@ -2659,6 +2715,20 @@ def _queue_tick():scene_ref = bpy.context.scenearch_mode = getattr(scene_ref, 'architecture_mode', 'sdxl')+ # Wait for any in-flight checkpoint / LoRA refreshes to finish so+ # ``model_name`` resolves to the correct value. The load_post+ # handler triggers an async refresh when the cached architecture+ # doesn't match the file that was just opened.+ from . import _pending_refreshes+ if _pending_refreshes > 0:+ if _queue_refresh_deadline == 0.0:+ _queue_refresh_deadline = _time.monotonic() + 15.0+ if _time.monotonic() < _queue_refresh_deadline:+ return 1.0 # Re-check on next tick+ else:+ print("[Queue] Model-list refresh timed out — proceeding anyway")+ _queue_refresh_deadline = 0.0 # Reset for next item+# Clear the error flag before invokingscene_ref.sg_last_gen_error = False@@ -2781,8 +2851,50 @@ def _queue_tick():from .render_tools import ExportOrbitGIFif ExportOrbitGIF._rendering:return 1.0 # Still rendering frames- # GIF export finished — save and advance+ # First GIF export finishedprint("[Queue] GIF export completed")++ # Check whether a second (no-PBR) export is needed+ wm = bpy.context.window_manager+ scene_ref = bpy.context.scene+ if (getattr(wm, 'sg_queue_gif_also_no_pbr', False)+ and getattr(scene_ref, 'pbr_decomposition', False)):+ # Start the no-PBR reproject → second GIF pipeline+ if _queue_start_no_pbr_reproject():+ _queue_phase = 'reprojecting_no_pbr'+ print("[Queue] Starting no-PBR reproject for second GIF")+ _tag_redraw()+ return 1.0+ else:+ print("[Queue] No-PBR reproject failed to start — skipping second GIF")++ _queue_save_and_advance(item)+ return 0.5++ # ── Phase: REPROJECTING_NO_PBR → wait for ComfyUIGenerate to finish ──+ if _queue_phase == 'reprojecting_no_pbr':+ from .generator import ComfyUIGenerate+ if ComfyUIGenerate._is_running:+ return 1.0+ # Reproject done — start second GIF with 1 sample + suffix+ print("[Queue] No-PBR reproject finished — starting second GIF export")+ if _queue_start_gif_export(samples_override=1, suffix='_no_pbr'):+ _queue_phase = 'exporting_gif_no_pbr'+ _tag_redraw()+ return 1.0+ else:+ print("[Queue] Second GIF export failed to start — restoring PBR")+ _queue_restore_pbr()+ _queue_save_and_advance(item)+ return 0.5++ # ── Phase: EXPORTING_GIF_NO_PBR → poll second GIF completion ──+ if _queue_phase == 'exporting_gif_no_pbr':+ from .render_tools import ExportOrbitGIF+ if ExportOrbitGIF._rendering:+ return 1.0+ print("[Queue] Second (no-PBR) GIF export completed")+ _queue_restore_pbr()_queue_save_and_advance(item)return 0.5@@ -2847,9 +2959,13 @@ def _queue_finish_item(item):_queue_save_and_advance(item)-def _queue_start_gif_export():+def _queue_start_gif_export(samples_override=None, suffix=''):"""Try to invoke ExportOrbitGIF with current WM settings.+ Args:+ samples_override: If set, use this sample count instead of the WM setting.+ suffix: Appended to orbit filenames (e.g. '_no_pbr' → orbit_no_pbr.gif).+Returns True if the operator was successfully started."""from .render_tools import ExportOrbitGIF@@ -2906,14 +3022,15 @@ def _queue_start_gif_export():'duration': getattr(wm, 'sg_queue_gif_duration', 5.0),'frame_rate': getattr(wm, 'sg_queue_gif_fps', 24),'resolution_percentage': getattr(wm, 'sg_queue_gif_resolution', 50),- 'samples': getattr(wm, 'sg_queue_gif_samples', 32),+ 'samples': samples_override if samples_override is not None else getattr(wm, 'sg_queue_gif_samples', 32),'engine': getattr(wm, 'sg_queue_gif_engine', 'CYCLES'),'interpolation': getattr(wm, 'sg_queue_gif_interpolation', 'LINEAR'),'use_hdri': use_hdri,'hdri_rotation': getattr(wm, 'sg_queue_gif_hdri_rotation', 0.0),'env_mode': getattr(wm, 'sg_queue_gif_env_mode', 'FIXED'),- 'use_shadow_plane': getattr(wm, 'sg_queue_gif_shadow_plane', False),'use_denoiser': getattr(wm, 'sg_queue_gif_denoiser', True),+ 'use_gpu': getattr(wm, 'sg_queue_gif_use_gpu', True),+ 'filename_suffix': suffix,}# Build a temp_override so the modal operator has a proper window@@ -2940,6 +3057,68 @@ def _queue_start_gif_export():return False+def _queue_start_no_pbr_reproject():+ """Disable PBR, then invoke a project-only reproject.++ Saves the original PBR / generation settings so they can be restored+ by ``_queue_restore_pbr()`` after the second GIF is done.++ Returns True if the reproject operator was successfully started.+ """+ global _queue_original_pbr, _queue_original_gen_method, _queue_original_overwrite++ scene = bpy.context.scene+ _queue_original_pbr = scene.pbr_decomposition+ _queue_original_gen_method = scene.generation_method+ _queue_original_overwrite = scene.overwrite_material++ scene.pbr_decomposition = False+ scene.generation_mode = 'project_only'+ scene.generation_method = 'separate'+ scene.overwrite_material = True++ # Select all cameras (required by ComfyUIGenerate)+ bpy.ops.object.select_all(action='DESELECT')+ for obj in scene.objects:+ if obj.type == 'CAMERA':+ obj.select_set(True)++ window = bpy.context.window+ wm = bpy.context.window_manager+ if not window:+ for w in wm.windows:+ window = w+ break+ if not window:+ print("[Queue] No window available — cannot start reproject")+ return False++ try:+ with bpy.context.temp_override(window=window):+ bpy.ops.object.test_stable('INVOKE_DEFAULT')+ print("[Queue] No-PBR reproject invoked")+ return True+ except Exception as e:+ print(f"[Queue] Failed to invoke reproject: {e}")+ import traceback; traceback.print_exc()+ # Restore settings on failure+ scene.pbr_decomposition = _queue_original_pbr+ scene.generation_mode = 'standard'+ scene.generation_method = _queue_original_gen_method+ scene.overwrite_material = _queue_original_overwrite+ return False+++def _queue_restore_pbr():+ """Restore PBR and generation settings saved by ``_queue_start_no_pbr_reproject``."""+ scene = bpy.context.scene+ scene.pbr_decomposition = _queue_original_pbr+ scene.generation_mode = 'standard'+ scene.generation_method = _queue_original_gen_method+ scene.overwrite_material = _queue_original_overwrite+ print(f"[Queue] Restored pbr_decomposition={_queue_original_pbr}")++def _queue_save_and_advance(item):"""Save the current .blend and advance to the next queue item."""global _queue_current_idx, _queue_phase, _queue_processing, _queue_timerdiff --git a/stablegen/utils.py b/stablegen/utils.pyindex f558708..5642b3b 100644--- a/stablegen/utils.py+++ b/stablegen/utils.py@@ -2,6 +2,29 @@ import bpyimport osfrom datetime import datetime+# ── Heavy modal operators that should block most other panel buttons ──+_SG_BLOCKED_MODALS = frozenset({+ 'OBJECT_OT_test_stable',+ 'OBJECT_OT_bake_textures',+ 'OBJECT_OT_add_cameras',+ 'OBJECT_OT_collect_camera_prompts',+ 'OBJECT_OT_trellis2_generate',+ 'OBJECT_OT_stablegen_reproject',+ 'OBJECT_OT_stablegen_regenerate',+ 'OBJECT_OT_stablegen_mirror_reproject',+ 'OBJECT_OT_export_orbit_gif',+})+++def sg_modal_active(context):+ """Return True if any StableGen heavy modal operator is currently running."""+ for window in context.window_manager.windows:+ for op in window.modal_operators:+ if op.bl_idname in _SG_BLOCKED_MODALS:+ return True+ return False++class AddHDRI(bpy.types.Operator):"""Add HDRI Global Illumination to the scene."""bl_idname = "object.add_hdri"@@ -23,6 +46,10 @@ class AddHDRI(bpy.types.Operator):max=10.0) # type: ignore+ @classmethod+ def poll(cls, context):+ return not sg_modal_active(context)+def execute(self, context):world = context.scene.worldif not world:@@ -100,8 +127,7 @@ class ApplyModifiers(bpy.types.Operator):@classmethoddef poll(cls, context):- # Operator can only run in Object Mode- return context.mode == 'OBJECT'+ return context.mode == 'OBJECT' and not sg_modal_active(context)def execute(self, context):if context.mode != 'OBJECT':@@ -184,6 +210,10 @@ class CurvesToMesh(bpy.types.Operator):bl_label = "Convert Curves to Mesh"bl_options = {'REGISTER', 'UNDO'}+ @classmethod+ def poll(cls, context):+ return not sg_modal_active(context)+def execute(self, context):# Select all objectsbpy.ops.object.select_all(action='SELECT')diff --git a/stablegen/workflows.py b/stablegen/workflows.pyindex b716290..b90e739 100644--- a/stablegen/workflows.py+++ b/stablegen/workflows.py@@ -20,12 +20,18 @@ from PIL import Imagedef _texturing_prompt(scene):- """Return the texturing portion of the prompt (after ``||`` if present).+ """Return the texturing portion of the prompt.- Users can write ``generation prompt || texturing prompt`` to give a- different prompt to the per-camera texturing step. When the separator- is absent the full prompt is returned unchanged.+ Priority order:+ 1. Dedicated ``texture_prompt`` field (if the separate-prompt toggle is on+ and the field is non-empty).+ 2. Legacy ``||`` separator in the main prompt (text after ``||``).+ 3. The full main prompt as-is."""+ if getattr(scene, 'use_separate_texture_prompt', False):+ tex = getattr(scene, 'texture_prompt', '').strip()+ if tex:+ return texraw = scene.comfyui_promptif '||' in raw:return raw.split('||', 1)[1].strip()@@ -33,11 +39,15 @@ def _texturing_prompt(scene):def _generation_prompt(scene):- """Return the generation portion of the prompt (before ``||`` if present).+ """Return the generation / mesh-creation portion of the prompt.- See :func:`_texturing_prompt` for the separator convention.+ When the separate-texture-prompt toggle is on the entire main prompt is+ treated as the generation prompt (no ``||`` splitting). Otherwise the+ legacy ``||`` separator is respected."""raw = scene.comfyui_prompt+ if getattr(scene, 'use_separate_texture_prompt', False):+ return raw.strip()if '||' in raw:return raw.split('||', 1)[0].strip()return raw@@ -1146,11 +1156,18 @@ class WorkflowManager:@staticmethoddef _cleanup_trellis2_temp_files():- """Remove TRELLIS2 IPC temp directories and voxelgrid cache files.-- The ComfyUI-TRELLIS2 custom node saves intermediate tensors and- voxelgrid data to disk but never deletes them, gradually filling- the system temp directory. This method cleans up both locations.+ """Remove stale TRELLIS2 IPC temp directories and voxelgrid cache.++ The ComfyUI-TRELLIS2 custom node creates one temp directory+ (``trellis2_<random>``) per worker subprocess via a module-level+ global and reuses it for the entire ComfyUI session. Deleting+ that directory while the worker is alive causes "Parent directory+ does not exist" errors on the next generation.++ Since only the most recently created directory can belong to the+ current (running) worker, we delete all *older* ``trellis2_*``+ directories entirely and only clean up the stale ``.pt`` tensor+ files inside the newest one, leaving the directory itself intact."""import tempfileimport glob@@ -1158,13 +1175,30 @@ class WorkflowManager:tmp_root = tempfile.gettempdir()- # IPC tensor directories: <tempdir>/trellis2_*/- for d in glob.glob(os.path.join(tmp_root, 'trellis2_*')):- if os.path.isdir(d):+ dirs = sorted(+ (d for d in glob.glob(os.path.join(tmp_root, 'trellis2_*'))+ if os.path.isdir(d)),+ key=os.path.getctime,+ )++ if not dirs:+ return++ # Everything except the newest is from a dead worker — remove entirely.+ for d in dirs[:-1]:+ try:+ shutil.rmtree(d)+ except OSError:+ pass++ # Newest directory may be in use — only purge .pt files inside it.+ newest = dirs[-1]+ for f in os.listdir(newest):+ if f.endswith('.pt'):try:- shutil.rmtree(d)+ os.remove(os.path.join(newest, f))except OSError:- pass # Still locked by worker — ignore+ pass# Voxelgrid cache (Windows: C:\tmp\trellis2_cache)for cache_dir in ('/tmp/trellis2_cache', r'C:\tmp\trellis2_cache'):@@ -2426,6 +2460,15 @@ class WorkflowManager:"On the ComfyUI host, delete C:\\Users\\<USERNAME>\\.triton\\cache ""and restart ComfyUI.")+ fail_on_exhausted = getattr(context.scene, 'trellis2_artifact_fail_on_exhausted', False)+ if fail_on_exhausted:+ error_msg = (+ "Mesh artifact check failed after all retries. "+ "Generation aborted (Fail if Unresolved is enabled)."+ )+ print(f"[TRELLIS2] ERROR: {error_msg}")+ self.operator._error = error_msg+ return {"error": error_msg}print(f"[TRELLIS2] WARNING: {warning_msg}")self.operator._warning = warning_msgreturn result@@ -2496,7 +2539,7 @@ class WorkflowManager:prompt[NODES['get_conditioning']]["inputs"]["image"] = [NODES['input_image'], 0]prompt[NODES['get_conditioning']]["inputs"]["mask"] = [NODES['input_image'], 1]else:- prompt[NODES['remove_bg']]["inputs"]["low_vram"] = scene.trellis2_low_vram+ prompt[NODES['remove_bg']]["inputs"]["low_vram"] = False# Configure conditioningprompt[NODES['get_conditioning']]["inputs"]["background_color"] = scene.trellis2_background_color