# Audio Dialog Isolation

Isolate vocals and accompaniment using a neural network (HT-Demucs FT, ONNX)

> Canonical page: https://elysiatools.com/en/tools/audio-dialog-isolation

- **Category:** Media

- **Keywords:** audio, dialog, isolation, vocal, stem, demucs, neural

## Overview

Runs an HT-Demucs FT neural network entirely in Node — no Python, no upload beyond the input file — and packages the separated vocals and accompaniment as a zip.

## Inputs

- **Audio File** (file): Select an audio file
- **Output Format** (select)

## When to use

- When you need to extract clean vocals or dialogue from a mixed audio track for remixing or voiceover work.
- When you want to remove vocals from a song to create a high-quality karaoke backing track or instrumental accompaniment.
- When you need to isolate speech from background music or noise in podcasts and video recordings.

## How it works

- Upload your audio file in any standard format up to the 100MB file size limit.
- Select your desired output format, such as WAV, FLAC, MP3, M4A, OGG, or Opus.
- The HT-Demucs FT neural network processes the audio to separate the vocal and accompaniment stems.
- Download the generated ZIP file containing the isolated audio tracks.

## Use cases

- Isolating dialogue from background music for video editing and post-production.
- Extracting clean vocal stems for music production, sampling, and song remixing.
- Creating instrumental backing tracks from full songs for karaoke or live performances.

## Frequently asked questions

### What audio formats are supported for upload?

You can upload any standard audio file, such as MP3, WAV, AAC, or FLAC, up to 100MB.

### Which output formats can I choose for the separated stems?

You can export the isolated tracks as WAV, FLAC, MP3, M4A, OGG, or Opus.

### How are the separated audio tracks delivered?

The tool packages the isolated vocal stem and the accompaniment stem together into a single ZIP file.

### Does this tool upload my audio to external servers?

No, the neural network runs locally in the application environment, ensuring your audio files remain secure.

### What neural network model does this tool use?

It uses the HT-Demucs FT model running via the ONNX runtime for high-fidelity source separation.

## Related tools

- [Batch Audio Rename](https://elysiatools.com/en/tools/audio-batch-rename): Batch rename audio files using patterns, text replacement, numbering, and case conversion. Returns renamed files as a ZIP download.
- [Audio Key Change for Singers](https://elysiatools.com/en/tools/audio-key-change-for-singers): Transpose a song by key or semitones while preserving its duration and speed.
- [Audio Stem Mixer](https://elysiatools.com/en/tools/audio-stem-mixer): Mix multiple stems and export vocal up/down versions
- [Audio Vocoder](https://elysiatools.com/en/tools/audio-vocoder): Apply a vocoder effect using a carrier and modulator
- [Audio Bitcrusher](https://elysiatools.com/en/tools/audio-bitcrusher): Reduce the sample rate and bit depth for a lo-fi effect
- [Audio Channel Swap](https://elysiatools.com/en/tools/audio-channel-swap): Swap the left and right channels of a stereo file
- [Audio Chipmunk Effect](https://elysiatools.com/en/tools/audio-chipmunk-effect): Increase pitch and speed for a chipmunk effect
- [Audio Denoise Chain](https://elysiatools.com/en/tools/audio-denoise-chain): Apply a multi-step denoise chain with filters and optional RNNoise

## Samples

- [Copyright-Free FLAC Audio Samples](https://elysiatools.com/en/samples/flac-samples): Lossless FLAC audio samples for testing and development, mirrored from MP3 set with nature sounds and meditation music
- [Copyright-Free MP3 Audio Samples](https://elysiatools.com/en/samples/mp3-samples): Collection of royalty-free audio samples for testing and development purposes including nature sounds, meditation music, and ambient audio
- [Copyright-Free WAV Audio Samples](https://elysiatools.com/en/samples/wav-samples): Uncompressed PCM WAV audio samples for testing and development, mirrored from MP3 set with nature sounds and meditation music
- [Copyright-Free Raw PCM Audio Samples](https://elysiatools.com/en/samples/pcm-samples): Raw PCM s16le audio samples featuring nature ambience and relaxing music for waveform, playback, and conversion workflows

## Related content

- [Karaoke and Stem Separation Tools](https://elysiatools.com/en/hubs/karaoke-and-stem-separation-tools): Separate vocals and accompaniment, clean the resulting stems, set a controlled balance, and prepare a karaoke or remix-ready delivery.