MIDILACE Runs on your device
TERM GUIDE · TUNING

Concert pitch: the note that sounds, referenced to A4 = 440 Hz.

Concert pitch is the pitch a note actually sounds at, as opposed to the pitch written on the page. The reference is A4 = 440 Hz, standardized as ISO 16. A MIDI file stores none of that: it stores note numbers. Number 69 means A4, and it only becomes 440 Hz when you apply the reference. On the site's measured file the five pitches are 57, 69, 72, 74 and 76 — 220.00, 440.00, 523.25, 587.33 and 659.26 Hz — and the spectrum of the same file puts the lowest tone at 219.8 Hz, which is 1.57 cents below nominal. Same note, different number: that gap is what concert pitch is about.

Convert a clip and read its hertz
THE DEFINITION

Sounding pitch versus written pitch

On a piano, written and sounding pitch are the same note — press the key labelled C and C sounds. On a transposing instrument they are not. A B♭ trumpet reading a written C sounds a B♭, a whole tone lower; an E♭ alto saxophone reading a written C sounds an E♭, a major sixth lower. The written note is a fingering instruction; the concert pitch is the result. When players from different instruments read the same score, concert pitch is the only language they share.

That distinction matters more, not less, once a file is involved. A MIDI note number is neither written nor sounding pitch — it is a position on a 128-step scale that becomes a frequency only when a reference is applied. The reference is a convention, not a property of the file.

THE CONVERSION

From note number to hertz, on the five pitches this file contains

One formula covers all 128 note numbers: f = 440 × 2(m − 69) / 12, where m is the MIDI note number. Every twelfth step doubles the frequency, which is why 57 is exactly half of 69 — an octave is a factor of two, and nothing else in the table is a round number.

The five pitches in the measured export, converted at A4 = 440 Hz
MIDI numberNote nameHertzWhere it appears in the file
57A3220.00seven events, 0.000–4.006 s — the sustained low lane
69A4440.00two events at 0.000 and 0.731 s — the octave double
72C5523.25two events at 0.986 and 1.742 s
74D5587.33one event at 2.961 s
76E5659.26one event at 2.009 s
Five MIDI note numbers converted to hertz: 57 is 220, 69 is 440, 72 is 523.25, 74 is 587.33, 76 is 659.26. A dashed marker shows the measured 219.8 hertz tone just under the nominal 220.
Note numbers on the left, hertz on the right, and the measured tone 0.2 Hz under its nominal value.
THE MEASURED GAP

219.8 Hz against 220 Hz is 1.57 cents — and that number changes with the reference

The spectrum of the 4.40-second file puts the sustained low tone at 219.8 Hz. The nominal value for A3 at A4 = 440 is 220.00 Hz, so the measured tone is 0.2 Hz low. Hertz differences mean little on their own; cents make them comparable, and one cent is one hundredth of a semitone:

1200 × log2(219.8 / 220) = −1.57 cents

Now change the reference and watch the same measurement move. If the file were read at A4 = 442 Hz — a reference some ensembles use — then A3 becomes 221.00 Hz, and the same 219.8 Hz tone sits at 1200 × log2(219.8 / 221) = −9.43 cents. Nothing about the audio changed. The reference did, and the tuning verdict changed with it. That is the whole practical content of "concert pitch": a frequency is a measurement, a note name is a convention, and the conversion between them needs the convention stated.

The same measured 219.8 Hz tone read against three references
ReferenceA3 in hertzMeasured tone in centsReads as
A4 = 440 Hz (ISO 16)220.00−1.57A3, essentially in tune
A4 = 442 Hz221.00−9.43A3, noticeably flat
A4 = 415 Hz (about a semitone down)207.50+99.7A♯3 / B♭3 — a different note name
WHY A FILE CANNOT ANSWER THIS

The export carries note numbers and no tuning reference

This is the part that surprises people: a MIDI file has no field for its reference pitch. It stores 57, and 57 is only 220 Hz by convention. Nothing in the file records whether the recording was tuned to 440, 442, or a guitar that drifted 8 cents sharp. The converter measured a frequency and rounded it to the nearest note number; the rounding is the loss.

So the honest reading of the export is two-layered: the note numbers are what the tool decided, and the hertz values are what those numbers mean under a stated reference. If you need the second layer to be exact — tuning a sample, matching a synth, checking whether a take was in tune — you have to measure the audio, not read the file. The measured 219.8 Hz above is that measurement, and it is the only place in this chain where a real frequency appears.

QUICK ANSWERS

Questions about concert pitch

Is concert pitch always 440 Hz?

440 Hz is the ISO 16 reference for A4, and it is the default in most software. It is a convention rather than a law: change the reference and every note number maps to a different frequency, as the table above shows. What stays constant is the ratio between notes, not their hertz. Every note at that reference is tabulated in the note frequency chart.

Why is A4 the reference and not C4?

Because A4 sits in the middle of the range most instruments share and is easy to tune to. The formula is written around it, so 69 is the anchor and every other note is a power of two away. Choosing C4 as the anchor would change the constant, not the intervals.

Does a converter output concert pitch or written pitch?

Concert pitch — it reports the frequency it heard, as a note number. If the source was a B♭ trumpet playing a written C, the export says B♭, because B♭ is what sounded. Transposition back to written pitch is a notation step done afterwards.

How far off can a note be before it is the wrong note?

Half a semitone, by definition: note numbers are 100 cents apart, so anything within 50 cents rounds to the nearer one. The 1.57-cent gap measured here is far inside that window, which is why 219.8 Hz and 220 Hz are the same note. What the gap tells you is tuning, not identity.

Can I check a file's tuning without the audio?

No. The file has already discarded the frequency when it rounded to a note number. You need the recording, or a spectrum of it, to see whether the take sat at 219.8 or 220.0 — and that is a different measurement from reading the notes.