VoiceStudio is the open-source, fully-local ElevenLabs alternative — voice cloning, voice design, video dubbing, dictation, transcription & audiobook creation in 646 languages.
Python aiaudiobookcudadubbingelevenlabs-alternativehuggingfacelocal-firstmlxomnivoice-studiospeech-to-texttext-to-speechtranscriptiontranslatettsvoice-aivoice-cloningvoice-generationvoicestudioworkflow
import { useState } from 'react';
import { LayoutGrid, Users, UserRound, Fingerprint } from 'lucide-react';
import SearchableSelect from '../SearchableSelect';
import { PRESET_ICONS, FALLBACK_VOICE_ICON, stripVoiceEmoji } from '../../utils/voiceIcons';
import toast from 'react-hot-toast';
import { PRESETS } from '../../utils/constants';
import { autoProfileId, assignSpeakerProfile, castParts, castSpeakers } from '../../utils/segments';
// The Default voice IS the empty profile_id, but a dataTransfer payload can't
// carry '' distinguishably from "no data" — so it travels as this sentinel.
const DEFAULT_VOICE = '__default__';
/**
* CAST — per-speaker voice assignment: the compact <select> strip
* (pre-existing, unchanged markup) plus an expandable casting board —
* draggable voice chips dropped onto speaker rows, with a keyboard path
* (focus a speaker, pick from a listbox). Both views write through
* `assignSpeakerProfile`, so a drop is byte-identical to picking the same
* option in the dropdown — no new persistence, the job save path already carries these fields.
*/
export default function CastingBoard({ t, dubSegments, setDubSegments, speakerClones, profiles }) {
const [boardOpen, setBoardOpen] = useState(false);
const [dropTarget, setDropTarget] = useState(null); // speaker id under a drag
const speakers = castSpeakers(dubSegments);
if (!speakers.length) return null;
const currentVoice = (spk) =>
dubSegments.find((s) => s.speaker_id === spk)?.profile_id ||
dubSegments.flatMap(castParts).find((part) => part.speaker_id === spk)?.profile_id ||
'';
const voiceLabel = (val, spk) => {
if (!val) return t('dub.default');
const clone = speakerClones[spk];
if (clone && val === autoProfileId(spk)) {
return stripVoiceEmoji(t('dub.from_video', { duration: clone.duration.toFixed(1) }));
}
if (val === autoProfileId(spk)) return `${t('dub.auto')} · ${spk}`;
if (val.startsWith('preset:')) {
return stripVoiceEmoji(t(`clone.preset_${val.replace('preset:', '')}`));
}
return stripVoiceEmoji(profiles.find((p) => p.id === val)?.name || val);
};
const paletteChips = [
{ value: '', label: t('dub.default'), Icon: UserRound },
...profiles.map((p) => ({
value: p.id,
label: stripVoiceEmoji(p.name),
group: 'clones',
groupLabel: t('dub.clone_profiles'),
Icon: Fingerprint,
})),
...PRESETS.map((p) => ({
value: `preset:${p.id}`,
label: stripVoiceEmoji(t(`clone.preset_${p.id}`)),
group: 'presets',
groupLabel: t('dub.design_presets'),
Icon: PRESET_ICONS[p.id] || FALLBACK_VOICE_ICON,
})),
];
// One option source keeps the compact selector, board, and drag validation in sync.
const voiceOptions = (spk) => [
...(speakerClones[spk] || currentVoice(spk) === autoProfileId(spk)
? [{ value: autoProfileId(spk), label: voiceLabel(autoProfileId(spk), spk), Icon: UserRound }]
: []),
...paletteChips,
];
const renderVoice = (option) => {
const Icon = option.Icon;
return (
<span className="inline-flex min-w-0 items-center gap-2">
<Icon size={16} aria-hidden="true" className="shrink-0 opacity-75" />
<span>{option.label}</span>
</span>
);
};
const assign = (spk, val) => {
setDubSegments(assignSpeakerProfile(dubSegments, spk, val));
toast.success(t('dub.casting_assigned', { voice: voiceLabel(val, spk), speaker: spk }));
};
const onDrop = (e, spk) => {
e.preventDefault();
setDropTarget(null);
const raw = e.dataTransfer?.getData('text/plain');
if (!raw) return;
const value = raw === DEFAULT_VOICE ? '' : raw;
if (!voiceOptions(spk).some((option) => option.value === value)) return;
assign(spk, value);
};
return (
<div className="dub-casting-controls my-3 p-4 bg-[var(--chrome-hover-bg)] rounded-lg border-0">
<div className="flex gap-[var(--space-2)] items-center flex-wrap">
<span
className="font-[family-name:var(--chrome-font-mono)] text-[length:var(--chrome-label-size)] text-[var(--chrome-fg-muted)] tracking-[var(--chrome-label-track)] uppercase font-semibold"
title={t('dub.cast_title')}
>
<Users size={16} aria-hidden="true" className="inline-block mr-2 align-text-bottom" />
{t('dub.cast')}
</span>
{speakers.map((spk) => {
return (
<div key={spk} className="dub-cast__pair">
<span className="font-[family-name:var(--chrome-font-mono)] text-[0.62rem] text-[var(--chrome-fg)]">
<UserRound
size={14}
aria-hidden="true"
className="inline-block mr-1 align-text-bottom"
/>
{spk}
</span>
<SearchableSelect
ariaLabel={spk}
menuPortal
renderGroupHeaders
buttonClassName="dub-cast__select min-h-10 rounded-lg border-0 bg-[var(--chrome-hover-bg)] px-3 py-2 text-sm text-[var(--chrome-fg)] hover:bg-[var(--chrome-accent-bg)] focus-visible:outline-2 focus-visible:outline-[var(--chrome-accent)]"
value={currentVoice(spk)}
onChange={(value) => setDubSegments(assignSpeakerProfile(dubSegments, spk, value))}
options={voiceOptions(spk)}
renderOption={renderVoice}
/>
</div>
);
})}
<button
type="button"
className={`dub-cast__board-toggle ${boardOpen ? 'is-open' : ''}`}
aria-expanded={boardOpen}
onClick={() => setBoardOpen((o) => !o)}
title={t('dub.casting_board_hint')}
>
<LayoutGrid size={15} aria-hidden="true" /> {t('dub.casting_board')}
</button>
</div>
{boardOpen && (
<div className="dub-cast__board" data-testid="casting-board">
<div className="dub-cast__palette" role="list" aria-label={t('dub.casting_voices')}>
{paletteChips.map((chip) => (
<span
key={chip.value || DEFAULT_VOICE}
role="listitem"
className="dub-cast__chip"
draggable
data-profile={chip.value}
title={chip.groupLabel ? `${chip.groupLabel} · ${chip.label}` : chip.label}
onDragStart={(e) => {
e.dataTransfer.setData('text/plain', chip.value || DEFAULT_VOICE);
e.dataTransfer.effectAllowed = 'copy';
}}
>
<chip.Icon size={18} aria-hidden="true" />
<span className="dub-cast__chip-text">
<span>{chip.label}</span>
{chip.groupLabel && <small>{chip.groupLabel}</small>}
</span>
</span>
))}
</div>
<p className="dub-cast__hint">{t('dub.casting_drag_hint')}</p>
<div className="dub-cast__rows">
{speakers.map((spk) => {
const clone = speakerClones[spk];
const options = voiceOptions(spk);
const current = currentVoice(spk);
return (
<div
key={spk}
className={`dub-cast__row ${dropTarget === spk ? 'is-drop' : ''}`}
data-speaker={spk}
onDragOver={(e) => {
e.preventDefault();
e.dataTransfer.dropEffect = 'copy';
}}
onDragEnter={() => setDropTarget(spk)}
onDragLeave={(e) => {
if (!e.relatedTarget || !e.currentTarget.contains(e.relatedTarget)) {
setDropTarget((cur) => (cur === spk ? null : cur));
}
}}
onDrop={(e) => onDrop(e, spk)}
>
<span className="dub-cast__row-speaker">
<UserRound size={16} aria-hidden="true" />
{spk}
</span>
<SearchableSelect
ariaLabel={t('dub.casting_assign_to', { speaker: spk })}
menuPortal
renderGroupHeaders
buttonClassName="dub-cast__row-select"
value={current}
options={options}
onChange={(value) => assign(spk, value)}
renderOption={renderVoice}
/>
{clone && (
<span
className="dub-cast__auto-chip"
title={t('dub.from_video', { duration: clone.duration.toFixed(1) })}
>
<UserRound size={13} aria-hidden="true" />
{stripVoiceEmoji(
t('dub.from_video', { duration: clone.duration.toFixed(1) }),
)}
</span>
)}
</div>
);
})}
</div>
</div>
)}
</div>
);
}