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 { afterEach, describe, expect, it, vi } from 'vitest';
import { decodeToMonoLowRate } from '../utils/audioTrim.js';
// #2558: decoding must not depend on a media element's metadata event, which a
// source may never emit; the Web Audio decoder is the only authority.
describe('decodeToMonoLowRate', () => {
afterEach(() => {
vi.unstubAllGlobals();
});
function stubDecoder(decode) {
const contexts = [];
class FakeOffline {
constructor(channels, length, sampleRate) {
contexts.push({ channels, length, sampleRate });
}
decodeAudioData(arr) {
return decode(arr);
}
}
vi.stubGlobal('OfflineAudioContext', FakeOffline);
// A media element that never reports metadata or an error.
vi.stubGlobal(
'Audio',
class {
addEventListener() {}
set src(_v) {}
},
);
return contexts;
}
it('decodes without waiting for media metadata', async () => {
const decoded = { duration: 3, sampleRate: 16000 };
const contexts = stubDecoder(async (arr) => {
expect(arr.byteLength).toBe(4);
return decoded;
});
await expect(decodeToMonoLowRate(new Blob([new Uint8Array(4)]), 16000)).resolves.toBe(decoded);
expect(contexts).toEqual([{ channels: 1, length: 1, sampleRate: 16000 }]);
});
it('propagates a decoder failure', async () => {
stubDecoder(async () => {
throw new Error('EncodingError');
});
await expect(decodeToMonoLowRate(new Blob([new Uint8Array(4)]))).rejects.toThrow(
'EncodingError',
);
});
});