saturalabs

Free tools for your next audio edit

Audio channel splitter. Hear each side on its own.

Split a stereo MP3 or WAV into separate left and right mono files. Prepare both together, preview each side, and download the files individually as WAV or re-encoded MP3. You can also choose one channel or an averaged mono mix. Your audio stays in this browser.

No account neededNo audio uploadWorks in your browser
Understand channels before separating sounds
Channel splitterFree · no sign-up
Download format

Try a 3-second stereo sample first:

Original synthesized tones, not AI results. Nothing plays automatically.

Your audio is read locally; this tool does not upload it. Export is 48 kHz mono: 16-bit PCM WAV, or MP3 at the selected bitrate. Original tags are not copied; neither format restores detail lost in the source. Channel extraction keeps existing audio; it does not separate voices or instruments mixed on the same channel.

Inspect the channels you already have

How to split stereo audio into mono files

Keep your original stereo file. A channel name describes where samples are stored, not which person or instrument they contain.

  1. 01 / OPEN

    Choose a stereo MP3 or WAV

    Choose one of the three built-in stereo samples to try the tool, or open your own two-channel recording up to 50 MiB and 5 minutes. The tool decodes it locally and shows a separate waveform overview for each channel. A mono file has no second channel to extract.

  2. 02 / CHOOSE

    Prepare both channels or one side

    Leave Both channels selected to prepare separate left and right mono files together. Choose WAV or MP3, play each labeled preview and download each file with its own button. If you only need one side, select Left channel or Right channel instead.

  3. 03 / COMPARE

    Keep the output your edit needs

    Download the same actual file you previewed. Choose WAV when both channels must share the same decoded start and duration, so place them at the same timeline position in your editor. For a single combined track, test Mono mix as well. Listen for voices or instruments becoming quieter or disappearing; an average is not always the right choice.

Existing channels, separate downloads

Left and right are channels. Vocals and drums are sources.

A stereo file stores two channels. They might hold different microphones, a backing track and a guide voice, or two sides of the same musical arrangement. Channel extraction keeps one of those stored channels; it does not estimate which sound belongs to a speaker or instrument.

For example, if an interview recorder placed microphone A on the left and microphone B on the right, separate mono files give you control over those recordings in an editor. Each microphone can still contain room noise and the other speaker. Selecting its channel does not remove that spill.

If a song’s voice and instruments are already mixed on both sides, neither side is automatically an acapella or instrumental. Use the vocal removal workflow for an estimated instrumental, or the stem separation workflow for a described source. Those are different operations from this free channel tool.

For desktop editing, Audacity also documents splitting stereo into separate mono tracks. In this tool, each exported file has one channel. When played through ordinary stereo equipment, its audio can be heard from both sides rather than retaining the original hard-left or hard-right position.

Choose by listening

Which output should you keep?

A useful recording on one side

Choose the channel containing the wanted recording. If the other side is empty or unwanted, averaging the two can make the wanted signal quieter without adding useful audio. Check the whole passage before discarding a channel.

Two microphones or speakers

Download both channels separately and place them at the same start time in your editor. Adjust each track as needed. This tool preserves the decoded duration and does not trim silence or choose a speaker for you.

A mono version of the complete stereo recording

Test Mono mix. It averages both channels with equal weights. If your project needs the stereo placement, retain the original instead. A one-channel WAV cannot preserve left-to-right positioning.

A predictable mix, with a listening check

How this stereo-to-mono mix works

For each sample frame, Mono mix uses the average below. It follows the stereo-to-mono rule in the Web Audio specification.

mono sample = (left sample + right sample) / 2

Identical left and right signals retain their sample amplitude after averaging. If the right side is silent, the left signal is halved. Equal signals with opposite polarity cancel. Real recordings may change only some sounds or frequencies, so a louder individual channel is a useful reference when checking the mix.

The tool does not normalize levels, align microphones, reverse polarity, repair clipping, or judge mono compatibility. Listen to the preview before choosing the mix. If it loses a wanted sound, keep an individual channel or use an editor with the controls needed to investigate the recording.

A different use of two channels

Try center cancellation on a stereo recording

Choose Center cancellation after opening your file. It subtracts the right channel from the left and halves the difference. Audio that is identical in both channels cancels; audio that differs can remain. The half scale leaves room for full-scale opposing signals. The result is one mono WAV or MP3, with its own real preview and download.

difference sample = (left sample − right sample) / 2

A centered lead vocal may be reduced if its samples match on both sides. Centered bass, kick, dialogue, and other wanted sounds can cancel too. Stereo reverb, doubled vocals, or a singer panned away from center may remain. This is arithmetic on channels, not a model that recognizes a singer.

For a predictable first test, open Matching channels above and choose Center cancellation: the identical synthesized tones should cancel to silence. A silent result from that example demonstrates matching-channel cancellation, not successful vocal removal. The Opposite polarity example instead reinforces in the difference, while Left sound, right silence becomes half-amplitude audio.

Audacity’s retained center-removal documentation explains classic channel subtraction and its limitations. Its older effect is no longer bundled with current Audacity; this page implements the explicit formula above and does not depend on that plugin. Listen to the remaining music as carefully as any reduced vocal. If wanted detail disappears, keep your original and compare a different method.

Know what you are exporting

Choose a mono WAV or MP3 for your next step

One channel, full decoded duration

Each WAV output is 48 kHz, 16-bit PCM. MP3 is re-encoded at 128, 192 or 320 kbps per mono file and may add codec delay or padding. Use WAV for precise timeline alignment. There is no edge fade or automatic silence trimming. The browser may resample the input, and compressed MP3 decoding is not a bit-for-bit extraction of the original file. Sample values above full scale are clipped to the PCM range; adjust excessive gain in an editor before exporting.

File size remains visible

Every single-file option has the same number of prepared mono samples, so their WAV byte sizes match. Actual MP3 byte sizes appear after encoding; mono at a specified bitrate does not mean half that bitrate. A mono WAV can still exceed the size of a compressed stereo MP3. Use the file size calculator to compare PCM and bitrate-based estimates.

Save both files before switching

Both channels provides two labeled previews and separate download buttons. Save each file you need; changing the selected output, format or bitrate clears the prepared results. Each new attempt uses the selected original, rather than re-encoding a previous result. Reloading the page clears this session; original tags are not copied and no recording is saved for later by the tool.

Troubleshoot one-sided recordings

Audio only plays in one ear?

Check the file and playback setup, then keep the recorded channel without averaging it with an empty side.

Compare before choosing an output

Read the whole-clip RMS comparison

After opening stereo audio, the tool reports RMS for the left channel, right channel, averaged mono signal and center-cancellation signal. Each measurement uses every decoded sample, not the sampled waveform overview. RMS is the square root of the mean squared sample amplitude; the displayed level is 20 × log₁₀(RMS), relative to amplitude 1. Exact digital silence is shown as −∞ rather than a fabricated floor.

Matching channels retain their RMS when averaged. If the right channel is silent, the mono mix has half the left channel's amplitude, about 6.02 dB lower than that active channel. Equal opposing samples cancel to silence. Center cancellation instead measures (left − right) / 2: matching channels cancel and equal opposing channels reinforce to the original channel amplitude. A low reading does not prove a vocal was removed; any matching sound can cancel. Try the built-in examples to compare these cases, then listen to the prepared file.

These readings describe the full decoded clip before PCM rounding, clipping or MP3 encoding. They are not integrated LUFS, true peak, a phase spectrum or proof that a mix is mono compatible. A short cancellation or missing instrument can be hidden by energy elsewhere. Keep the original and inspect the difficult passage by ear. The dB to gain calculator explains how a level difference relates to an amplitude ratio.

Listen to three channel examples

Hear when mixing to mono keeps, weakens, or cancels a sound

Compare original stereo, separate left and right channels, and averaged mono. Download a source WAV to try it in this free splitter.

MP3 encoder credits

Built with the LAME encoder via lamejs.

The optional MP3 download uses lamejs 1.2.1 by Alex Zhukov, a JavaScript port of the LAME MP3 encoder. Its unmodified script is served separately. See the attribution and source notes, upstream license notice, LGPL-3.0 and GPL-3.0 texts, and complete upstream source package.

A few useful answers

Free audio channel splitter questions

Is the audio channel splitter free?

Yes. Extracting a channel, preparing a mono mix, previewing, and downloading either WAV or MP3 are free and require no account. They happen in this browser. Saturalabs’ separate AI isolation service uses its own account and download credits.

How do I split a stereo file into left and right audio?

Open a stereo MP3 or WAV and select Both channels. Choose WAV or MP3, prepare the files, then use the separate left and right preview and download controls. Both mono WAVs keep the same decoded start and full duration; MP3 can include codec delay or padding. You can also prepare just one side or an averaged mono mix. No ZIP or automatic multiple-file download is created.

Can I download separate channels as MP3?

Yes. Choose MP3 and 128, 192 or 320 kbps; each requested mono file is encoded once at that bitrate. Both channels prepares two individual MP3s with labeled previews and download buttons. A new lossy encode can discard detail and add delay or padding, so use WAV for precise editing. Actual file sizes appear after export, and neither format restores information missing from the original.

Can this separate vocals from instruments?

Selecting left or right keeps every sound already on that side. The separate Center cancellation option subtracts matching stereo content and can reduce centered vocals, but also removes centered instruments. It is not a stem separator and cannot recognize a singer. Compare an AI separation workflow when channel arithmetic is unsuitable.

What does the mono mix option do?

It averages corresponding left and right samples: (left + right) / 2. It does not choose the cleaner side, normalize loudness, or repair phase differences. Opposing signals can cancel, so compare the mono mix with both individual channels before using it.

What is the difference between Mono mix and Center cancellation?

Mono mix uses (left + right) / 2, so identical channels remain. Center cancellation uses (left − right) / 2, so identical channels cancel to silence. Both produce one-channel WAV or MP3 files; neither identifies voices or instruments. Listen before choosing, because wanted sounds can disappear.

Which files and export settings are supported?

Use a stereo MP3 or WAV up to 50 MiB and 5 minutes. Browser codec support varies; standard PCM WAV is a useful fallback. Each output is one-channel, 48 kHz: 16-bit PCM WAV or MP3 at 128, 192 or 320 kbps. The browser may resample the input; MP3 is a new lossy encode and may add delay or padding. Original tags and compressed bytes are not preserved.

Why is my mono file rejected?

A mono file has only one stored channel, so it has no separate left and right audio to extract. Playing mono through two speakers does not add a second recorded channel. Use the free trimmer if you simply need a shorter mono excerpt.

After preparing your audio

Focus on the sound inside it.