FLAC to MIDI: check the input before you trust the output.
Start with a decodable lossless file, a clear solo melody, and a control recording. This page turns the site's measured FLAC run into an input checklist; it does not claim full-song accuracy.
Input conditions
Use the public source.flac fixture and keep the workflow local in the browser. The tested file was decoded to mono 11025 Hz PCM before the same detector used for the MP3 comparison.
The page accepts FLAC alongside MP3, PCM WAV, OGG, and M4A. AIFF is not supported in this browser workflow.
What the measured run returned
| Input | Output | Observed result |
|---|---|---|
| FLAC lossless | flac-lossless.mid | Four events with pitches 62, 65, 69, 74 |
| Silent FLAC | silent-flac.mid | Zero events |
Those values belong to this isolated four-note fixture. They are not a universal FLAC accuracy score.
Failure checks
If decoding fails, keep the original file and try a standard PCM WAV or MP3 only as a separate comparison. If the input is silence, the measured control returned zero events. If it is a dense mix, chords, drums, or layered vocals, record that as an untested condition instead of assuming the fixture result applies.
Questions before conversion
Does a FLAC file always produce better MIDI?
Not from this test alone. The measured FLAC and 128 kbps MP3 comparison returned the same four events and timings for the isolated melody.
What should I do when the sample rate differs?
Keep the file and detector settings documented, then compare the decoded result with the source. Only the tested mono 11025 Hz decode is covered by this record.
Where are the reproducible records?
Use the source file, run script, detector, and results JSON in the public evidence folder. The results record is the reference for this page.