Why "video" and not just MP4
Files arrive in whatever their source produced. An iPhone gives you MOV. A download gives you MKV. A browser screen recorder gives you WebM. An old camcorder tape capture gives you AVI, and a broadcast recording gives you TS. All of them are containers, and all of them wrap an audio stream that can be pulled out.
That is the whole reason to convert video to MP3 through something that reads every container rather than one that only knows MP4. The step that has to work is demuxing, and it is different for each format. Whatever you started with, the job is the same: extract audio from video, then encode it.
What lives inside each one
The container tells you very little about the audio codec. Roughly:
- MP4 and MOV almost always hold AAC
- WebM holds Opus, or Vorbis on older files
- MKV holds anything at all, including lossless FLAC
- AVI usually holds MP3 or AC-3
- TS from a broadcast typically holds AC-3 or AAC
This matters for one practical reason. If your MKV happens to contain FLAC, then converting it to MP3 is a genuine first-generation encode from lossless material, and 320 kbps will sound excellent. If your WebM contains a 96 kbps Opus stream, no MP3 bitrate can add back what Opus already discarded, and picking 320 just makes a bigger file.
Sample rate, channels and metadata
The source sample rate is kept as it is. Nothing is resampled to 48 kHz, and a mono source stays mono instead of being doubled into two identical channels. Titles and artist tags come across when the container declares them.
Files up to 500 MB are accepted, which covers most single clips. Beyond that a browser tab cannot hold the container, so the tool says so before starting rather than dying partway.