In this guide
For your own upload, start with the original video or a downloaded copy. To record permitted playback, select a loopback or app-audio source, make a short test and check the saved file. Both methods preserve the mixed soundtrack rather than separating voice from music.
To extract audio from a YouTube video you uploaded, use the original video file if you have it. Otherwise, download your own upload from YouTube Studio when available, then convert the video file to an audio format such as MP3 or WAV. Saturalabs' free video-to-audio converter accepts a supported local MP4 or MOV; it does not download a YouTube URL.
Looking for the title instead of an audio file? See how to find a song from a video.
For someone else's video, ask the creator for an authorized source file and the rights needed for your intended use. A video being publicly viewable does not itself give you a separate audio file or permission to reuse its soundtrack.
To record permitted YouTube playback instead, capture the computer output with an appropriate recorder, test the saved audio and export the result. On Windows, Audacity can use WASAPI loopback; on a supported Mac, OBS can capture application audio in a video recording. Recording happens in real time and can include other sounds sent to the selected output.
Record YouTube playback on Windows with Audacity
- Choose the output. Start a video you have permission to record and note which speakers or headphones are playing it. Disable unrelated notification sounds; loopback may capture them too.
- Select loopback. In Audacity, open Audio Setup → Host and choose Windows WASAPI. Under the recording device, choose the loopback entry for the same output device. Audacity's desktop-audio instructions explain this routing.
- Make a short test. With playback active, press Record, then stop after a few seconds and play the recording. If it is silent, verify the output and loopback device match. Turn off audible input monitoring to avoid feedback or doubled sound.
- Capture and export. Start a fresh recording for the needed passage, stop it, listen through the result and use File → Export Audio to save WAV or MP3. Preserve an unedited copy and check any silent gaps: loopback behavior can affect timing when playback stops.
Record playback on a Mac with OBS
On macOS 13 or later, OBS Studio 30 and later provides a macOS Audio Capture source that can capture desktop or application audio. Add that source to a scene, select the browser or desktop as appropriate, allow requested permissions, and confirm the mixer meter moves while a permitted video plays. Record a short test and play the saved file. OBS describes supported versions and capture options in its Mac desktop-audio guide.
OBS normally saves a video recording rather than a standalone MP3. If the saved file is a supported MP4 or MOV, the video-to-audio converter can prepare a short WAV or MP3 locally within its limits. For a longer recording or a different container, use a desktop editor that can open it. See recording computer audio for source, feedback and troubleshooting details.
Playback capture does not create permission to reuse the audio. Check the video's terms and obtain the rights needed for your purpose. For your own upload, the original file or an available YouTube Studio download can avoid a real-time recording and may provide a better starting point.
Choose the right source and output
| Your goal | Start here | Check before use |
|---|---|---|
| Edit the complete soundtrack | Original project audio or a WAV export from your video | Sync, sample rate and whether all wanted sounds are present. |
| Listen to a smaller audio file | Convert a supported local video to MP3 | Bitrate, playback and the additional lossy encoding. |
| Keep voice without music | Use a separate original voice track or test voice isolation | Music bleed and changes to words; extraction alone keeps the mix. |
1. Find your original or download your own upload
The original export or editing project is often the best starting point: it may contain separate tracks and avoids an extra round trip through an online video copy. If you need a copy of your upload, YouTube's instructions say to sign in to YouTube Studio, open Content, find the video, and choose Download from its menu. Availability and output resolution depend on the video; YouTube lists restrictions, including some removed or restricted uploads.
Google also offers Takeout to export your YouTube data. An archive can take time, and a Brand Account may require switching accounts. YouTube's offline viewing downloads are intended for playback in its app and are not a standalone MP3 export.
2. Convert a short video to MP3 or WAV
- Open the file. Visit the free converter and select an MP4 or MOV from your device. It reads the file locally without an account or video upload.
- Choose a format. WAV creates 16-bit PCM at 44.1 or 48 kHz; MP3 is re-encoded at 48 kHz with 128, 192 or 320 kbps options. WAV is a practical working copy for editing; MP3 is lossy and can add delay or padding.
- Convert, preview and download. Listen to the actual output and keep the original video. Check timing against picture before returning audio to an edit.
The file and output must each fit 50 MiB and the audio must be no longer than five minutes. The decoded audio must be mono or stereo, and browser codec support varies. The browser chooses the stream: this tool cannot select among multiple audio tracks. For a longer file, stream selection or copying the encoded audio without a new encode, use a desktop editor or the desktop extraction methods.
Why does the file still contain music and speech?
Extracting the soundtrack moves the existing mix into a separate file; it does not unmix it. If you have the editing project, export the original voice or music track separately. Otherwise, try the voice isolator on a representative passage and compare the isolated result with the original. Separation is an estimate and can leave bleed or alter speech. This AI workflow is separate from the free converter and requires an account and sufficient credits before isolation starts; its download dialog shows the cost.
How to extract music from a YouTube video
First decide whether you want the complete soundtrack or the music without other sounds. If the video already contains only the music you need, converting an authorized local copy to MP3 or WAV retains that mix. If speech or sound effects overlap the music, soundtrack extraction keeps those sounds too.
- Find an authorized local source. Use your original project or video, an available download of your own upload, or a file supplied by the owner with permission for your intended use. Saturalabs does not accept a YouTube link.
- Check for a separate music track. If the original project has one, export it directly. If you only have a finished mixed recording, select a supported MP4, MOV, MP3 or WAV up to 50 MiB in the background music extractor.
- Describe the music and audition it. Request the audible background music and review Isolated sound for the estimated music. Background audio is the remainder; it is the wrong output if your goal is to keep the requested music.
- Check before downloading. Compare the beginning, a passage with overlapping speech, and the end against the original. Listen for missing instruments, damaged transients and voice bleed. WAV downloads require an account and sufficient credits; the result is audio, not a new video.
An isolated music estimate is not necessarily the original song or a studio-quality stem. If singing is part of the music you want to retain, listen carefully for lost vocals; if you instead want an instrumental backing, the separate vocal-removal workflow below targets the singer and uses the remainder. See the music extraction guide for a more detailed comparison.
How to remove vocals from a YouTube video
To make a backing track from a video you have permission to use, start with its original instrumental or separate project tracks if available. If you only have a mixed soundtrack, use an authorized local copy of the video and test vocal separation. Saturalabs does not take a YouTube URL.
- Choose the recording. Open the vocal remover and select a supported MP4, MOV, MP3 or WAV no larger than 50 MiB. You can upload the video directly; there is no need to extract its soundtrack first.
- Target the singing. Describe “The singing vocals.” Listen to Background audio for the remaining music. Isolated sound is the requested vocal, useful for comparison but the opposite choice for a backing track.
- Review before using. Compare a verse, busy chorus and ending with the original. Listen for remaining lyrics, backing harmonies, lost instruments and vocal reverb. Downloads are WAV audio and require an account and sufficient credits; the result does not include an edited video.
- Return it to your video if needed. Import the chosen audio into a video editor, align it against the picture and mute the original soundtrack to avoid playing both versions. Check the start, end and sync; keep a copy of your original project.
Voice removal can also remove wanted speech or change instruments that overlap singing. When you need only music, a clean original instrumental is preferable if available. If you want to keep speech while reducing background music, the video music-removal guide describes the different target and output choice.
Common questions
Can I paste a YouTube link into the converter?
No. Choose a supported video file stored on your device. Use your original, an available copy of your upload, or a file supplied by its owner with the necessary permission.
Is WAV automatically higher quality than the original soundtrack?
No. Converting compressed audio to WAV cannot restore discarded detail. WAV avoids another lossy encode in this step, but the source and earlier processing still determine what is available.
Why will my file not convert?
Check that it contains audio and fits the size and duration limits. MP4 and MOV can contain different codecs; a browser may not decode every file. Try the original file, another compatible browser or a desktop editor.