The CoMSAr Database architecture addresses the requirement for sustainable and interoperable access strategies to institutional collections of sound- and audio-visual recordings by providing the technological concepts and operational structures required to facilitate the global networking of digital sound archives e.g.

The UHH Ethnographic Sound Recordings Archive


as well as efficient, decentralized online access to their holdings.

The CoMSAr framework allows for comprehensive annotations and meta-data management as well as for analytical content description, classification, indexing, evaluation, and retrieval of sound files by employing a common and consistent set of parametric descriptors derived directly from the audio signal.

The GUI (Graphical User Interface) of the web application provides an intuitive entry point for database interaction. Users may retrieve assets by keyword or browse the collection visually using the Self-Organizing feature map. The GUI adapts to different screen sizes and various devices by implementing concepts of Responsive Web Design, such as CSS media queries, a grid system and flexible media and images. No further software needs to be installed by clients. 

The backbone of the COMSAR database's server is provided by the LAMP Stack (Linux operating system, Apache HTTP Server, MySQL relational database management system and, PHP programming language).

The back end provides capabilities for multi-level user and permission management, full-text search as well as an extendable set of database operations to manage and access archival media assets and according annotations and meta-data. Furthermore, format-agnostic handling of media files has been implemented to support a wide range of audio and video file formats such as AVI, WMV, WAV, MP4, MP3, QuickTime, OGG and FLAC. Integrity of relations between media file and meta-data is ensured by utilizing MD5 hashes.

Application Programming Interface (API)

The system’s API implements the public Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH) and a Dublin Core compatible API endpoint, based on Dublin Core Element Set, Version 1.1 and Dublin Core Metadata Terms. The endpoint serves XML data following the IASA recommendations for digital audio objects. It is furthermore designed to share contents across multiple archive instances. These technologies open the field for computational interaction with peer archives and meta data harvesters. It allows contribution to digital cultural heritage platforms such as Europeana, Dariah, Dismarc, and

Public OAI-PMH endpoint (RESTful XML API) for UHH Ethnographic Sound Recordings Archive E.S.R.A.



1st May