Resources← Back to the neck

Resources / Tunings / Tuning — why the sources disagree

Tuning — why the sources disagree

The problem

There is no single answer to "how is a cümbüş tuned." Research during the prototype build turned up at least three mutually incompatible answers, all from sources that present themselves as authoritative.

This is not a gap in the research. It's a property of the instrument.

The three positions

1. Turkish reference works: all fourths. The cümbüş is given as Do–Sol–Re–La–Mi–Si read from the thinnest string — C5 B... i.e. B2 E3 A3 D4 G4 C5 low to high. Every interval a perfect fourth. Sources note this sits a fourth above the classical Turkish oud's Sol–Re–La–Mi–Si–Fa♯. Attested

2. Tuner applications: oud layouts. Turkish ud/tanbur/cümbüş tuner apps ship two cümbüş presets — a "pes" (low) and a "tiz" (high) — and both are oud tunings, not the all-fourths one. Attested

3. Players: whatever the piece needs. A player on a Turkish forum gave the practical rule: the cümbüş uses the oud's tuning, and the bam (thickest course) is retuned to the karar of whatever piece is being played. [attested, and consistent with practice on related instruments]

Why position 3 is probably the real answer

The bam is a drone/bass course, not a melody course. Its job is to sound under the tonic. If the tonic changes, the bam moves. This is the same logic that governs scordatura on any instrument with a designated bass string, and it explains why the "standard" disagrees between sources — they're each capturing a snapshot of a variable.

Consequence for the app: do not present one tuning as canonical. Present the families, explain the bam convention, and let the user retune.

The E2–A2–B2–E3–A3–D4 problem

The most widely circulated "Turkish cümbüş tuning" online is E2 A2 B2 E3 A3 D4 — intervals 4th, 2nd, 4th, 4th, 4th.

Uncertain I could not match this to a primary source. It is internally coherent (it's the Turkish oud layout with a low E added), plenty of players use it, and it was the prototype's default. But it should carry a lower confidence marker than the ones with citations, and the app should say so rather than presenting it as settled.

This is exactly the kind of claim that hardens into fact through repetition. Mark it.

Why documentation is thin

Attested A Turkish player put it bluntly on a forum: there is essentially nowhere teaching cümbüş, no established method book exists because the instrument is young, and in practice oud players play it using oud technique.

That's the whole situation. The cümbüş was invented in the 20th century, it never acquired a conservatory pedagogy, and its literature is borrowed wholesale from the oud.

This is good news for a learner. Anything written for oud transfers directly, because the tuning is the same family and the left hand is doing the same work. Learn from oud material; apply to cümbüş.

Octave assignments are the weakest data

Pitch classes are well-attested. Specific octaves are inferred, and they genuinely vary with scale length, string gauge, and how hard the player is willing to push a skin head.

Practice Many players tune the whole instrument down a semitone or a whole tone for comfort and to spare the head. Nothing wrong with it — transpose everything and carry on.

The data model should mark octave assignments as inferred rather than asserting them. See 00-project/data-model.md.

Sources

Meshk uses privacy-respecting analytics (Google Analytics, proxied through this domain) to see which tunings, makams, and features people actually use. Read the privacy policy.