Section for Speech and Audio Analysis

General information

CSL 4400The Section for Speech and Audio Analysis currently employs six specialists with a linguistic or engineering background. Three of the staff have forensic expert status, one of whom has a PhD degree.

Head of Section:

Agata Trawińska, MA

The scope of expert examinations encompasses:

  1. Examination of the authenticity of recordings (analogue and digital), i.e. verifying:
    • whether an analogue recording is continuous
    • whether an analogue recording is original
    • whether several recordings were made using the same device
    • whether a seized device was used to record evidence material
    • whether a digital recording is authentic; verifying the date and the time when the recording was made
  2. Speaker recognition:
    • identifying whether different recordings contain the utterances of the same speaker
    • comparative speaker identification using reference material that has been recorded in the Institute or submitted for examination
  3. Transcriptions of the recording content:
    • transcription of low quality recording
    • identifying the number of speakers
  4. Others activities, e.g. recording enhancement in order to improve speech intelligibility, creating copies of recordings made on untypical devices, technical support in voice line-up.

Wizualizator MOSES2

The Section for Speech and Audio Analysis studios are equipped with:

  • active midfield monitor loudspeakers in two studios;
  • four workstations with open and closed dynamic headphones;
  • a package for digital edition and enhancement of audio recordings;
  • digital recorders with microphones with various characteristics;
  • software for acoustic parameters extraction and numerical data processing;
  • magneto-optical visualization system for magnetic audio tapes.

Selected publications:

  1. Trawińska A., Kajstura M., The inbuilt recorder of mobile phones – possibilities of forensic speaker identification, Problems of Forensic Sciences 2004, 57, 51–80.
  2. Kajstura M., Trawińska A., Hebenstreit J., Application of the electrical network frequency (ENF) criterion. A case of digital recording, Forensic Science International 2005, 155, 165–171.
  3. Klus K., Trawińska A., Forensic speaker identification by the linguistic-acoustic method in KEÚ and IES, Problems of Forensic Sciences 2009, 78, 160–174.
  4. Brzozowski J., Kajstura M., Determination of optimal order of performance of fingerprint and speech and audio analysis on magnetic tapes, Problems of Forensic 2009 (in press).
  5. Michałek M ., Application of the electrical network frequency criterion in the authenticity examination of digital recordings, Problems of Forensic Sciences (in press).

The Section is involved in the research work of the Expert Working Group for Forensic Speech and Audio Analysis of the European Network of Forensic Science Institutes and verifies experts’ competence in proficiency tests drawn up by the group. Staff also give lectures and run training courses for law enforcement professionals as well as promoting forensic research among students.