Most of the Arabic scales in actual use contain a note that does not exist on a keyboard — not a note you are playing wrong, a note with no key. This is the single biggest reason a programmed "Arabic" melody sounds like an impression of one.
Written by M III D O · 19 September 2026
Jins Hijaz — the four notes behind the sound everyone calls Arabic — is built from these intervals, in tones: ½, 1½, ½. Every one of those exists in twelve-tone tuning. It plays perfectly on a piano.
Now here are two ajnas that turn up far more often in Egyptian repertoire than hijaz does:
½ 1½ ½
Plays on a keyboard. No microtones anywhere.1 ¾ ¾ 1
Two three-quarter tones. Does not exist on a keyboard.¾ ¾ 1
Opens on a three-quarter tone. Same problem, immediately.Interval sizes as notated by Maqam World. A "1" is a whole tone, "½" a semitone.
A three-quarter tone is one and a half semitones. Halfway between the black key and the white key. There is no such key, there is no such MIDI note, and there is no rounding that gets you there — round it down and you have a minor scale, round it up and you have a major one. Both are wrong, and both are what most plugins do.
Twelve semitones in an octave, and the note Rast needs sits between two of them:
The third degree is not E and it is not E♭. It is between them, and it is the note the whole maqam is recognised by. Play E and the line sounds major and cheerful; play E♭ and it sounds minor and sad. Rast is neither, and that in-between quality is the entire point.
You will read that Arabic music divides the octave into 24 equal quarter tones. That is a writing system, not a tuning system. It exists so the music can be put on paper at all — every note is rounded to the nearest quarter tone so a stave can hold it.
The actual intonation is finer than that and it varies. It varies by maqam, by region, by which note you are approaching from, and by player. A performer in Cairo and one in Aleppo will place the same written note differently, and both are correct. It is carried by ear and by teaching, not by a table of frequencies.
So the honest statement is not "Arabic music uses quarter tones". It is: Arabic music uses intervals that twelve-tone tuning has no name for, and the quarter-tone notation is the compromise invented to write them down.
In 1932 the first Congress of Arab Music met in Cairo — the major institutional attempt to put Arab music theory on a formal footing, with delegations from across the Arab world and Europe. One of its committees took on exactly this question: should the octave be fixed at 24 equal steps?
The Egyptian delegation argued for it — a standard grid, writable, teachable, tunable on instruments. The Turkish delegation, represented by Rauf Yekta Bey, and the Syro-Lebanese delegates argued against: equal division would flatten the differences their traditions were actually made of.
The congress ended without agreement. Egypt convened two further symposia, in 1959 and 1964, to try to settle the same dispute.
This is worth sitting with. The question "how many notes are in an Arabic scale" was put to the best musicians and scholars of the tradition, in a formal setting, and they could not agree — because the thing being measured was never a grid in the first place. Your DAW is not failing at an easy problem. It is applying a grid to music that resisted one on the record.
The most honest answer. A recorded oud, ney, qanun or voice brings the intonation with it, because it was played rather than placed. This is why loop packs recorded with real players sit differently from ones drawn in — the microtones are in the audio whether or not anyone writing the pack could notate them.
On a monophonic lead you can bend the offending degree down by 50 cents and hold it. It works, it is fiddly, and it breaks the moment you want two notes at once. Fine for a ney line, useless for a chord.
Any sampler or synth that accepts a tuning file lets you move a single scale degree by a fixed number of cents and leave the rest alone. That is the closest a piano roll gets to being honest: you are not writing microtonal MIDI, you are telling the instrument that its E is not where it thinks.
Hijaz has no microtones. Nahawand has none. Kurd has none. If you are working in a western arrangement with chords underneath, these are not a consolation prize — they are the maqamat that genuinely survive the trip. Writing well in one of them beats writing badly in Rast.
What is not honest is programming Rast, rounding its third to E♭, and calling the result Arabic. That is a minor scale. It is a perfectly good minor scale. It is just not the thing it is being sold as.
The interval structures of jins rast, bayati and hijaz are read from the notation published by Maqam World, the standard online reference for Arabic maqam theory. The account of the 1932 Congress of Arab Music — the split between the Egyptian and the Turkish and Syro-Lebanese positions on 24 equal steps, the failure to reach agreement, and the later Egyptian symposia of 1959 and 1964 — follows the published record of the congress, including the documented summary of its committees and the research now built on its recordings, such as the open dataset of the 1932 recordings. What to do in a DAW is mine.