MP4 to MIDI

Which audio track does this converter use?

We tested that question on the live site in desktop Chrome on September 23, 2026. We generated two small MP4 files with a black picture and two separate AAC audio tracks. One track contained a constant A4 tone at 440 Hz; the other contained A5 at 880 Hz. These were synthetic diagnostics, not recordings of a musical performance.

In the first file, A4 was the first audio track and A5 was second. In the second file, we swapped the two audio streams without re-encoding them. We marked the second audio track as the default in both files. This made track order and the default flag different signals, so the result could show which one the converter followed.

Both conversions completed and produced downloadable MIDI files. The source videos were generated for two seconds; the converter displayed 2.04 seconds of decoded audio for each. We report the displayed duration separately instead of assuming it must match the video duration exactly.

We kept the default detection settings for both runs: Lowest note C3, Highest note C6, Detection confidence 0.85, Quiet-sound filter −45 dB, Shortest note 100 ms and Transpose 0. Swapping the stream order changed the detected pitch from A4 to A5. In these samples, the converter followed the first audio track even though the second track carried the default flag.

Repeat the check or troubleshoot the wrong soundtrack

1. Download either test MP4 above, open the converter and select it with Choose audio or video. Wait until the status says the audio has loaded. The first use may need to download the local media decoder before this happens.
2. Listen to the input audio player on the converter page. That player previews the decoded track used for detection. It is more relevant to this conversion than a different player's choice of soundtrack. The site has no dropdown for choosing another audio track.
3. Leave the listed settings unchanged, convert and inspect the displayed note range. Repeat with the other test file. Use Preview MIDI for the synthesized output and download the result after reviewing it.
4. If your own video's intended melody is on another audio track, prepare a copy containing that track in your video or audio editor before selecting it here. Merely changing the default-track flag is not the fix demonstrated by this test. The converter does not expose an alternate-track selector.
5. If the correct soundtrack already contains several instruments mixed together, selecting an audio track does not separate those instruments. Use a solo recording or an isolated melody you already have, rather than expecting this control path to perform stem separation.

For comparison, download our actual A4 MIDI result and A5 MIDI result. The fixture record contains the generation commands, stream order and file hashes. The video codec in these examples is MPEG-4 Part 2 and the audio codec is AAC. Successful conversion of these files does not establish support for every codec an MP4 container might carry.