Contributions are governed by a memorandum of understanding, tailored to each contributor's conditions. The commitments below are the project's standing principles; specific scope, attribution format and any embargo are agreed with each contributor before any data changes hands.

Attribution is permanent and machine-safe. The person or body who made a recording is credited at every stage. Credit is bound to the audio through file-naming conventions and embedded metadata, and is preserved through machine processing — automated sorting, classification and machine-learning steps operate on stems that retain their provenance, so a recording can always be traced back to its source.

Originals are never altered. Each contributed master is checksummed on receipt and kept unmodified. All segmentation, enhancement and analysis is performed on derivative copies; the original remains intact and retrievable.

Every edit is declared. Each sound carries a complete metadata record — including latitude and longitude, recording equipment, the recordist and their purpose, and species. Any post-production or audio-repair applied by the project is logged in that record, so an edited file is never mistaken for a raw one and the scientific chain of custody stays transparent.

Not-for-profit, defined use. Recordings are used for research, conservation and listening — never sold or licensed for commercial profit. Contributors set the terms of access, may set embargoes until their own work is published, and retain the right to correct attribution.

Detailed terms and MoU templates will be published here.