On July 22, researchers at the Music Intelligence Lab at the American University of Beirut released the Digital Arabic Maqām Archive (DiArMaqAr) — a multilingual, open-source platform, corpus, and API for Arabic maqām theory — together with Tanghīm, a free open-source MIDI tool that applies those tunings live in a DAW. The project is the work of Dr. Khyam Allami and Ibrahim El Khansa, and the archive is available in English, Arabic, and French at diarmaqar.net, with the code on GitHub.
What’s in the archive
The corpus gathers hundreds of tuning systems, along with the ajnās (the small melodic units maqāmāt are assembled from) and the maqāmāt themselves. Crucially, it doesn’t force one representation: the same material can be viewed as string length, fret length, pitch class, staff notation, or cents, which matters enormously because those notations encode different traditions of describing the same music. A theorist working from historical instrument measurements and a producer working in cents are looking at the same object through incompatible lenses; the archive translates between them.
The API is the piece that makes this research infrastructure rather than a reference website. Tuning data that a program can query is tuning data that can end up inside instruments, analysis tools, notation software, and machine-learning pipelines — instead of being locked in PDFs and out-of-print treatises.
Tanghīm: live tuning, not preset tuning
The companion plugin, Tanghīm, runs on macOS, Windows, and Linux and works with Oddsound MTS-ESP, MPE, and monophonic pitch bend, with Max for Live devices for Ableton users. Its stated design goal is the interesting part: to function like tuning pegs inside your software — letting a performer adjust intonation while playing, rather than picking a fixed scale from a menu before the session starts.
That distinction is the whole argument. Most microtonal support in modern software is selection-based: load a Scala file, and the notes are where the file says they are for the rest of the session. But in maqām practice — as in many traditions — intonation is an expressive, situational act. The exact placement of a pitch shifts with the melodic phrase, the mood, the performer, the moment. Freeze it into a lookup table and you’ve preserved the coordinates while discarding the practice.
Why this belongs in a creative-technology feed
Digital music tools carry embedded assumptions, and the deepest one is 12-tone equal temperament: piano rolls, MIDI note numbers, quantized grids, and nearly every synth preset assume it. Anyone working outside that system spends their time fighting the tooling rather than making music with it.
Projects like this — Allami has previous form here, having co-created browser tools for exploring alternative tunings — are effectively infrastructure work for musical diversity. Making a tradition’s theory machine-readable and its intonation performable in standard production software lowers the cost of working outside the default from “build your own instrument” to “install a plugin.” It’s also a healthy corrective to how generative music models are trained: a corpus this structured is exactly the kind of resource that could keep the next generation of AI music tools from flattening everything into the same tempered grid.
Both the archive and the plugin are free and open source.