echogarden/dist/api/SourceSeparation.js

An easy-to-use speech toolset. Includes tools for synthesis, recognition, alignment, speech translation, language detection, source separation and more.

import { attenuateIfClippingInPlace, ensureRawAudio, subtractAudio } from '../audio/AudioUtilities.js';
import { Logger } from '../utilities/Logger.js';
import { extendDeep } from '../utilities/ObjectUtilities.js';
import { loadPackage } from '../utilities/PackageManager.js';
import chalk from 'chalk';
import { readdir } from '../utilities/FileSystem.js';
import { defaultMDXNetOptions, getProfileForMDXNetModelName } from '../source-separation/MDXNetSourceSeparation.js';
import { joinPath } from '../utilities/PathUtilities.js';
export async function isolate(input, options) {
    const logger = new Logger();
    const startTimestamp = logger.getTimestamp();
    const inputRawAudio = await ensureRawAudio(input);
    let isolatedRawAudio;
    let backgroundRawAudio;
    options = extendDeep(defaultSourceSeparationOptions, options);
    switch (options.engine) {
        case 'mdx-net': {
            const MDXNetSourceSeparation = await import('../source-separation/MDXNetSourceSeparation.js');
            const mdxNetOptions = options.mdxNet;
            const packageDir = await loadPackage(`mdxnet-${mdxNetOptions.model}`);
            const modelFilename = (await readdir(packageDir)).filter(name => name.endsWith('onnx'))[0];
            if (!modelFilename) {
                throw new Error(`Couldn't find an ONNX model file in package directory`);
            }
            const modelPath = joinPath(packageDir, modelFilename);
            await logger.startAsync(`Convert audio to 44.1 kHz stereo`);
            let inputRawAudioAs44100Stereo = await ensureRawAudio(inputRawAudio, 44100, 2);
            logger.end();
            const modelProfile = getProfileForMDXNetModelName(mdxNetOptions.model);
            isolatedRawAudio = await MDXNetSourceSeparation.isolate(inputRawAudioAs44100Stereo, modelPath, modelProfile, mdxNetOptions);
            logger.end();
            // Release memory for the converted input audio since it's not needed anymore
            inputRawAudioAs44100Stereo = undefined;
            await logger.startAsync(`Convert isolated audio back to original sample rate (${inputRawAudio.sampleRate} Hz) and channel count (${inputRawAudio.audioChannels.length})`);
            isolatedRawAudio = await ensureRawAudio(isolatedRawAudio, inputRawAudio.sampleRate, inputRawAudio.audioChannels.length);
            await logger.startAsync(`Subtract from original waveform to extract background audio`);
            backgroundRawAudio = subtractAudio(inputRawAudio, isolatedRawAudio);
            break;
        }
        default: {
            throw new Error(`Engine '${options.engine}' is not supported`);
        }
    }
    await logger.startAsync(`Postprocess audio`);
    attenuateIfClippingInPlace(isolatedRawAudio);
    attenuateIfClippingInPlace(backgroundRawAudio);
    logger.end();
    logger.logDuration(`Total source separation time`, startTimestamp, chalk.magentaBright);
    return {
        inputRawAudio,
        isolatedRawAudio,
        backgroundRawAudio
    };
}
export const defaultSourceSeparationOptions = {
    engine: 'mdx-net',
    mdxNet: defaultMDXNetOptions,
};
export const sourceSeparationEngines = [
    {
        id: 'mdx-net',
        name: 'MDX-NET',
        description: 'Deep learning audio source separation architecture by KUIELAB (Korea University).',
        type: 'local'
    },
];
//# sourceMappingURL=SourceSeparation.js.map