MIDI input check

Does DC offset affect MIDI conversion?

Not in this controlled eight-note fixture. DC offsets from 0.05 through 0.35 retained all eight notes, with zero median misses and extra events.

Five DC-offset levels from zero through 0.35 each retain eight MIDI notes.
The constant signal offset was the only changed condition.
Short answer: do not remove DC offset solely to rescue a clear short melody conversion. This fixture stayed at 8 of 8 retained notes through the tested 0.35 offset. Best with one exposed melody, not chords or a full mix.

Convert one clear melody

Controlled DC-offset check

Added DC offsetMedian across 10 fixtures
0, 0.05, 0.108 retained; 0 missed; 0 extra
0.20, 0.358 retained; 0 missed; 0 extra

What this does and does not rule out

DC offset is a constant vertical shift in a waveform, not a second pitch. This test kept the same C4/E4 phrase, note duration, harmonics, and level while changing only that constant shift. It does not show that clipping, noise, a competing instrument, or very quiet audio will behave the same way.

Fixture conditions and limits

Each point used 10 mono 44.1 kHz synthetic PCM fixtures with eight alternating C4/E4 voice-like harmonic notes, 120 ms per note, 0.70 source amplitude, and an 8 ms fade. Melody / solo instrument mode, Balanced sensitivity, a 2048-sample analysis window, and a 1024-sample hop stayed fixed. Retained means exactly one output overlaps at least 50% and is within 50 cents. Detector SHA-256: DDA185AC8FB6A179A4FCFA73FD4197ACC95B8B6F59B233F5D3E330F596309F9F.

These are controlled source checks, not a universal DC-offset or accuracy promise.

Related: audio clipping · quiet audio · audio-to-MIDI source checks