Audio Forensics Toolbox

Recording, editing and coding steps usually leave characteristic traces within audio material. The Audio Forensics Toolbox developed by Fraunhofer IDMT allows users to analyze such traces and detect the reuse of material. Broadcasters, media producers, and media archives may use the Toolbox to conduct quality assessments, facilitate more efficient annotation of metadata, and detect unintended or intended editing of audio material.

Some of the respective algorithms have been developed within the European project REWIND.


Reliable assessment of the quality of encoded audio material

Using the Audio Forensics technologies, it is possible to detect previous encoding steps and coding parameters in audio material. Using that, quality and redundancy problems can be avoided, e.g. by avoiding intended transcoding, and selecting appropriate codecs and coding bitrates for content processing and storage for a given context.

Easy creation and linking of metadata for content production, archiving, and search

The Audio Forensics technologies allow users to identify the devices and microphones used for a recording, to detect cuts and splices in audio material, and to find out whether certain segments have been reused in a new context. This can be used to automatically enrich metadata and support respective search capabilities for large media archives. The detection of reused audio segments allows automatic and efficient linking and tracking of metadata and copyright information, which helps avoid redundancies and inconsistencies.

Semi-automatic tampering detection for user-generated content

Whether presumably authentic audio material has actually been manipulated is a key question (e.g. for journalists researching a story). The Audio Forensics technologies are able to identify certain types of manipulation, helping users assess the authenticity of certain material.

Technical Information

The following methods are used to analyze the audio material:

  • Codec analysis (for AAC, MP3, MP3PRO, and GSM)
  • Microphone classification and discrimination
  • Electric network frequency (ENF) and stable tone analysis
  • Audio segment matching

Please contact us for an offer on your individual analysis case.

Learn more



Multi-modal, crowd-supported technology platform for disinformation analysis



Center of excellence for AI in media: Robust audio forensics, Privacy, Recommendation


Explainable AI-based disinformation and provenance analysis



Robust authentication of synthetic audio material



Detection of AI-based speech synthesis and voice alienation for law enforcement



Integration of audio forensic tools into the content verification platform TrulyMedia



REVerse engineering von audio-VIsual coNtent Data: audio footprint detection

Research topic

Media Forensics

Trustworthy media content