Catalog series

Middle Eastern Ensemble

Rights-cleared, articulation-level recordings of Middle Eastern strings, winds, and percussion, from oud and qanun to darbuka doum and tek, captured as controllable acoustic primitives for audio AI.

6 instruments Reference Sessions

Why this series matters

Middle Eastern music lives in the detail: microtonal intervals of the maqam system, breath-shaped ney tone, plucked and tremolo oud and qanun, and the doum, tek, and ka strokes of the darbuka. General-purpose training data rarely captures these in isolation. This series is designed to make those controls learnable, not merely present.

Maqam intervals

Stroke vocabulary

Articulation-level capture

Provenance-first

What makes it hard to capture

Performers fluent in maqam and traditional technique
Microtonal pitch that falls between Western equal-tempered notes
Breath, string, and jingle sources that need clean isolation
Metadata that reflects maqam and rhythmic-cycle logic, not generic tags

What the series includes

Articulation-level datasets for each instrument: isolated notes, ornaments, and techniques, individually labeled
Reference Sessions: ensemble performances delivered as dry stems, processed stems, and the finished mix, in traditional and contemporary arrangements

Typical use cases

Microtonal pitch modeling and maqam-aware generation
Stroke-level percussion classification and generation
Ornament and tremolo modeling for plucked strings
Evaluating how models handle non-Western tuning

Provenance and annotation

Every dataset is documented under the Proteus Standard™: source provenance for each file, written rights assignment from every performer, and per-file SHA-256 checksums with licensed data.

Orpheus annotation is available as an optional layer for teams that need human-aligned control vocabularies.

Previews and licensing

Each dataset starts with a free preview on Hugging Face. Full datasets are licensed at the Research, Startup, or Enterprise tier, and HFA also produces commissioned datasets to a buyer's specification.

Discuss licensing Browse datasets

Datasets in this series

Select a dataset to view its details and preview.

Darbuka

IN PRODUCTION

Isolated darbuka doum, tek, and ka strokes for modeling goblet drum articulation and rhythmic-cycle patterns.

View dataset →

Middle Eastern Ensemble

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique. Duis cursus, mi quis viverra ornare, eros dolor interdum nulla, ut commodo diam libero vitae erat. Aenean faucibus nibh et justo cursus id rutrum lorem imperdiet. Nunc ut sem vitae risus tristique posuere.

IN PRODUCTION

Ney

IN PRODUCTION

Isolated ney notes and breath articulations for modeling airy end-blown flute tone and microtonal inflection.

View dataset →

Middle Eastern Ensemble

IN PRODUCTION

Oud

IN PRODUCTION

Isolated oud notes, tremolo, and ornaments for modeling fretless plucked pitch, maqam intervals, and phrasing.

View dataset →

Middle Eastern Ensemble

IN PRODUCTION

Qanun

IN PRODUCTION

Isolated qanun notes and figures for modeling plucked zither tone, microtonal tuning, and ornamental runs.

View dataset →

Middle Eastern Ensemble

IN PRODUCTION

Riq

IN PRODUCTION

Isolated riq strokes and jingle articulations for modeling frame drum timbre and fine rhythmic ornamentation.

View dataset →

Middle Eastern Ensemble

IN PRODUCTION

Saz

IN PRODUCTION

Isolated saz notes and strummed figures for modeling long-necked lute tone, microtonal frets, and phrasing.

View dataset →

Middle Eastern Ensemble

IN PRODUCTION