toolgarden.xyz
中文
stem separationvocal isolationinstrumental extractorHT-Demucslocal processing

How to Split a Song into Vocals, Drums, Bass, Guitar, Piano, and Instrumental Stems

Learn how browser-local stem separation works, when to choose the four- or six-stem model, and how to export vocals, drums, bass, guitar, piano, and other parts as WAV files.

ToolGarden tools prioritize browser-local processing, so files and text do not need to be uploaded to a server.

Published July 30, 2026About 9 minutesBy ToolGarden

You do not always need a full desktop audio suite to extract vocals, make an instrumental, or focus on the drums, bass, guitar, or piano in a song. A browser music-source-separation model can analyze the mix and export estimated parts as separate WAV files.

ToolGarden uses an open-source HT-Demucs ONNX model. Audio decoding, model inference, overlap-add processing, and WAV encoding happen locally in the browser. The first run downloads model assets, but the selected song is not uploaded as a remote separation job.

What is audio stem separation?

A released song is usually a stereo mix in which vocals, drums, bass, and other instruments share the same waveform. It cannot be unpacked like a ZIP file to recover the original studio tracks. A separation model instead estimates how much each source contributes by analyzing spectral patterns, rhythm, timbre, and musical context.

The results can work well for practice, karaoke backing tracks, transcription, arrangement study, and mix references, but they are not identical to the original multitrack session. Reverb, distortion, layered performances, and overlapping frequencies can leave audible bleed between stems.

Should you choose four stems or six stems?

ModelOutputsBest forMain tradeoff
Four-stem HT-DemucsVocals, drums, bass, otherVocals, instrumentals, and rhythm practiceMore focused vocals and other, without separate guitar or piano
Six-stem HT-DemucsVocals, drums, bass, other, guitar, pianoGuitar, piano, or six-part analysisAdds two stems, while other becomes narrower and may sound weaker

Choose four stems when your goal is vocal isolation or an instrumental. Choose six only when guitar or piano must be available separately. More outputs do not automatically mean better overall quality; they divide the same mixture into smaller categories.

How to split a song in the browser

  1. Open the stem splitter, then click the upload area or drag a song onto the page.
  2. Choose the four- or six-stem model. Four stems is the better default for vocals and instrumentals.
  3. Select only the stems you need. Avoiding unnecessary long outputs reduces result memory use.
  4. Select “Separate stems” and keep the page open while it decodes audio, downloads and loads the model, separates segments, and encodes WAV files.
  5. Preview the results, download individual stems, or package every result in a ZIP archive.

Why is the first run slower?

The page is small, but the neural-network models are not. The current four-stem model is about 158 MiB and the six-stem model about 130 MiB. A model is downloaded only when separation begins, verified, and stored in the browser cache so the same browser can usually reuse it later.

After the download, the browser still needs to create an ONNX Runtime session and load the parameters, which can take 10–15 seconds. The smaller files store FP16 weights, but inference still needs more than 1 GB of working memory. Download size and runtime memory are different measurements.

  • Use a current desktop version of Chrome or Edge.
  • Close memory-heavy tabs and applications before processing.
  • Trim a long song to the section you need before separation.
  • Do not reload, suspend, or close the page while the model is loading or running.
  • Mobile browsers normally have stricter memory limits and are a poor fit for this model.

Does local processing mean no network traffic?

No. The first run needs a connection to download the page code, ONNX Runtime, and model file. The relevant privacy boundary is that the site downloads software and public model assets while the user-selected audio remains in the browser instead of being posted to a remote separation API.

Model, WASM, and Worker requests in the Network panel are expected. To verify a local-processing claim, inspect whether the selected song appears in a request body rather than treating every network request as an upload. Network use normally drops after the required assets are cached.

How should you evaluate the separated audio?

Do not listen to only one isolated track. Recombine vocals, drums, bass, and other to check whether they approximately reconstruct the source, then inspect the stem that matters most. Light metallic artifacts and bleed may be unavoidable; the practical question is whether they interfere with your intended use.

  • Karaoke: check for obvious lead-vocal residue and holes left in the instrumental.
  • Instrument practice: prioritize stable timing, melody, and continuity over studio-grade isolation.
  • Mixing or sampling: audition on headphones and speakers, and check phase, bass, and transient artifacts.
  • Transcription or analysis: use the stem with the clearest target voice or instrument.

Common failures and practical fixes

SymptomLikely causeWhat to try
Model download failsThe model host is blocked, the connection stopped, or cached bytes are incompleteCheck the connection and retry; clear the cached model if necessary
Model appears stuck while loadingThe browser is verifying the model and creating a local inference sessionAllow 10–15 seconds and avoid repeated clicks or reloads
Model cannot be loadedThe browser or device does not have enough available memoryClose other tabs and retry in desktop Chrome or Edge
Separation stops partway throughThe audio is too long, too many output stems were selected, or inference ran out of memoryTrim the audio, select fewer outputs, and run it again
A stem contains audible bleedSources overlap heavily in frequency, reverb, distortion, or stereo positionTry the other model and judge the result against the intended use

Copyright and responsible use

The ability to separate a recording does not automatically grant permission to publish, sell, perform, sample, or redistribute the results. Commercial releases, public videos, performances, and training datasets may involve rights in the recording, composition, and performance. Process audio you own, are authorized to use, or may lawfully use in your jurisdiction.

Key takeaways

The most important stem-separation choice is not “more stems is better.” Use four stems for vocals and instrumentals, and six only when guitar or piano must be isolated. Browser-local inference reduces the privacy exposure of uploading a song, but a large model still requires download time, substantial memory, and realistic expectations about separation quality.

Frequently asked questions

Q.Is my song uploaded during stem separation?

No. Audio decoding, inference, overlap-add processing, and WAV encoding happen locally in the browser. The first use downloads model and runtime assets, but downloading a model is different from uploading the selected song.

Q.Should I use four stems or six stems to make an instrumental?

Use four stems in most cases. It outputs vocals, drums, bass, and other, so you can omit vocals and recombine the instrumental parts. Choose six only when you also need guitar or piano separately.

Q.Why does a 130 MB model need more than 1 GB of memory?

The file size measures stored weights. Inference must expand parameters, create an ONNX Runtime session, allocate intermediate tensors, and hold output buffers, so working memory is much larger than the downloaded file.

Q.Why can I still hear instruments in the vocal stem?

Source separation estimates components of a finished mix rather than recovering original studio tracks. Shared frequencies, reverb, distortion, and stereo placement make some bleed difficult to avoid.

Q.Can I split stems on a phone?

Some high-end devices may succeed, but it is not recommended. Mobile browsers have tighter memory, background suspension, and thermal limits, so desktop Chrome or Edge is normally more reliable.