The short answer: no — not on its own. This converter is a pitch detector that reads from the whole signal, not a source separator, so there is no mode that hands you only the bass. If the bass is the only low content in the file (an isolated stem), it can be read — but only down to about 55 Hz and with coarser resolution below 181 Hz. We measured both cases on the converter’s own code, and the numbers are below. For what a full mix comes back as, that is Song to MIDI: what a full mix comes back as.
1. Why “just the bass” is not a button
Every file is resampled to 22 050 Hz, read in 2048-sample windows stepped 256 samples at a time, and searched for pitch inside a fixed band:
- F_MIN = 55 Hz — the lowest pitch the search will consider.
- F_MAX = 1800 Hz — the monophonic search ceiling.
- P_MAX = 4000 Hz — the highest spectral peak polyphonic mode keeps.
There is no stage in that pipeline that says “find the bass and ignore everything else.” The two modes behave differently, but neither isolates a voice:
- Monophonic follows the single strongest line it can track — one melody, not “the bass.”
- Polyphonic keeps up to several spectral peaks per frame, so it returns more of the mix, not a cleaner slice of one instrument.
The exported .mid is one channel with no program-change events and no track names — there is literally no field in it that could tag a note as “bass.” Separation has to happen before conversion, in a DAW. What is actually inside a MIDI file covers that empty program field.
2. Measurement 1: a mix returns everything, not the bass
We built a clip with a bass riff (A2–E2–A2–C2) playing under a sustained C4–E4–G4 chord at equal loudness, and ran it through the converter’s own analyze() in both modes:
| Mode | Distinct pitches returned | What that tells you |
|---|---|---|
| Polyphonic | 6 — C2, E2, A2, C4, E4, G4 | Bass and chord, all mixed into one list |
| Monophonic | 2 — C2, A2 | One line only, and it even missed the E2 bass note |
Neither mode gives you the bass alone. Polyphonic mode returned the bass and the chord together — nine notes, six distinct pitches, no way to tell which were the bass. Monophonic mode returned a single line (just A2 and C2), and it dropped the E2 entirely. So the question “can it pull out just the bass?” has no setting that answers yes: you either get the whole mix or one untagged line.
3. So how do you actually get the bass?
You isolate the bass yourself, then convert the isolated stem. The converter reads pitches; it does not do the separation:
- Take the bass stem, not the master. In your DAW, mute or solo the bass, or export just the bass track, then point the converter at that file.
- Convert that stem. With the bass as the only low content, monophonic mode will track it as one clean line. The FL Studio import, Ableton import and Logic Pro import pages cover getting the resulting
.midback into each. - Watch the low end. Even on a clean stem, the bass has a read floor (next section), so check the lowest notes survived.
This is the same shape as the drum problem: the engine is built to find pitch, not to unmix a recording. Mono vs poly changes which notes you get, never whether a voice is separated.
4. Measurement 2: the read floor on an isolated bass stem
We fed a single descending bass line — A2, E2, C2, A1, and E1 (the last at 41 Hz, below the 55 Hz floor) — as the only content, and read it in both modes:
| Mode | Notes caught | Notes missed |
|---|---|---|
| Monophonic | A1, C2, E2, A2 (down to 55 Hz) | E1 (41 Hz, below floor) |
| Polyphonic | C2, E2, A2 | A1 (55 Hz) and E1 (41 Hz) |
The practical floor is about 55 Hz. Monophonic mode caught the A1 (55 Hz) but dropped the E1 at 41 Hz, which sits under F_MIN and is invisible by construction. Polyphonic mode was worse here — it missed the A1 too, because a near-floor pure tone has weak harmonics for the peak finder to lock onto. The lesson: even an isolated bass won’t give you sub-bass. A 5-string bass’ low B (30 Hz) or an 808 sub will simply not appear.
5. Measurement 3: where the bass reads clean, and where it doesn’t
For contrast we fed a mid-range melody (C4–E4–G4–C5) through the same code. Both modes returned all four notes, 4 of 4. The bass line in section 4 did not:
| Material | Mono caught | Poly caught |
|---|---|---|
| Mid melody (C4–C5) | 4 of 4 | 4 of 4 |
| Bass stem (A2–E1) | 4 of 5 | 3 of 5 |
The gap is the resolution, not the detector being “for melody.” Below roughly 181 Hz (MIDI 53.6, about F#3) two neighbouring semitones are closer together than one 10.77 Hz analysis bin, so the low end is read coarse. That 181 Hz line is measured in Why MP3 loses notes that WAV keeps, where the same 2048-sample window produces it directly. Above it, notes separate cleanly; below it, they blur — which is exactly why the bass stem lost notes the mid melody kept.
6. How to check your own bass on this converter
The behaviour above repeats on your audio because the constants are fixed. Drop a bass-heavy file in and read the .csv it writes — header index,midi_note,note_name,start_sec,end_sec,duration_sec,velocity:
- Sort by note name. If your lowest bass note is below about A1 (55 Hz), expect it to be missing entirely.
- Run mono and poly back to back. On a busy mix, poly shows the bass buried among chord and melody notes; mono gives one line that may not even be the bass.
- Isolate first. Export the bass stem from your DAW, then convert that. The note list will read as one clean low line instead of a scramble.
- Compare to a mid melody. If a C4 phrase comes back 4 of 4 but your A1 phrase comes back 2 of 4, you are seeing the 181 Hz resolution drop, not a file problem.
Questions people actually ask
Can I get just the bass as a separate MIDI track?
No, not from a full mix in one step. The converter outputs one channel with no instrument tags, so a mixed file comes back as one undifferentiated note list. To get a bass-only track, isolate the bass stem in a DAW first, then convert that stem.
Will polyphonic mode separate the bass from the rest?
No. Polyphonic mode keeps more spectral peaks per frame, so it returns more of the mix — bass, chords and melody all together — not a cleaner slice of one voice. In our mixed test it returned six pitches at once (bass plus chord) with no way to tell them apart.
Why did my low E or drop-D note not show up?
Anything below about 55 Hz (A1) is outside the search band and is never detected. A low E on a 4-string bass is ~41 Hz and a drop-D is ~37 Hz, both under the floor. The note simply does not exist inside the analysis, regardless of sensitivity.
Does converting from WAV instead of MP3 help the bass?
It helps cleanliness, not the floor. A lossless file gives the detector a steadier harmonic structure, but the 55 Hz floor and the 181 Hz resolution drop come from the analysis window, which is identical for WAV and MP3. See Why MP3 loses notes that WAV keeps for the measured numbers.
What is the lowest bass note this converter can read?
About A1 (55 Hz) in monophonic mode on a clean stem; polyphonic mode may drop it. Below that — low B, low E, 808 subs — there is nothing to read.
Drop a file in and watch the piano roll — the Audio to MIDI Converter runs entirely in your browser, so every number above is yours to reproduce on your own bass in a few seconds. If the notes you get need tidying afterwards, How to clean up a converted MIDI before you use it walks the passes that turn a raw export into a usable part.
Related guides
- Song to MIDIWhat a full mix comes back as
- Mono vs polyWhat each detection mode misses
- MP3 to MIDIWhy MP3 loses notes that WAV keeps
- MIDI converterWhat is actually inside a MIDI file
- Drums to MIDIWhat a pitch detector does with no pitch to find
- FL Studio importThe drag that opens the .mid, and why there is no sound until you load an instrument
- Clean up a converted MIDIThe 5 passes that turn a raw export into a usable part