How to extract audio from a video (and save it as MP3)
By Muhammad Sohaib and Luqman Ahmad · NexixLabs · Updated
Sometimes the picture doesn’t matter: a recorded lecture, a podcast interview filmed for social media, a concert clip or a voice memo recorded as a video. Extracting the audio gives you a file that is a fraction of the size and plays on anything, from a car stereo to a smart speaker.
Here is how to do it well – and quickly – even with very long recordings.
What is inside a video file
Almost every video file contains a separate audio track, usually AAC in MP4 and MOV files, Opus in WebM, and anything from MP3 to AC3 in MKV and AVI. Extracting the audio means reading that track, ignoring the video and saving the sound in a new file.
Saving as MP3 involves re-encoding, because the original track is rarely MP3 already. That is fine: at a sensible bitrate the difference is inaudible, and MP3 is the most compatible audio format there is.
Choosing a bitrate
The bitrate decides the size and the quality of the MP3:
- 64–96 kbps: voice notes, phone calls and lectures where only speech matters.
- 128 kbps: clean speech and podcasts – the usual choice for spoken content.
- 192 kbps: music and anything with a wide range of sound.
How big will the MP3 be?
An MP3’s size is its bitrate times its length. At 128 kbps one minute is about 1 MB, so a one-hour lecture is roughly 55–60 MB. At 192 kbps the same hour is about 85 MB. Compared with a video of the same length – often several gigabytes – that is tiny.
Choosing a higher bitrate than the original audio does not improve the sound; it only makes the file larger. Most phone recordings have AAC audio around 128–256 kbps.
Very large and long recordings
Online converters usually limit file size, because the whole video has to be uploaded first – which can take longer than the conversion itself. A browser-based tool that works locally reads the file straight from your disk instead. With Asset Compressor, a long recording is never uploaded; the limit is only your device’s memory, and a computer copes with multi-gigabyte files far better than a phone.
For the fastest results with long files, trim first so only the part you need is processed.
Trim to the part you need
Under Advanced you can set a start and an end time, for example to keep a single song from a concert recording or one answer from an interview. Trimming before extraction keeps the MP3 small and saves you a trip to an audio editor.
Only use content you have the rights to
Extracting audio from your own recordings, lectures you are allowed to record, or material you have permission to use is fine. Asset Compressor does not download from YouTube or other websites – it only works with files already on your device – and you should not use it to copy music or videos you don’t have the rights to.