Transcription and voice comparison of noisy interceptions: Remarks from an audio forensics report

Sonia Cenceschi; Chiara Meluzzi

doi:10.17469/O2110AISV000007

Autori

Sonia Cenceschi Digital Forensic Service, Department of Innovative Technologies, University of Applied Sciences and Arts of Southern Switzerland, Switzerland https://orcid.org/0000-0002-4145-9593
Chiara Meluzzi Dipartimento di Studi Letterari, Filologici e Linguistici, Università degli Studi di Milano https://orcid.org/0000-0002-2291-006X

DOI:

https://doi.org/10.17469/O2110AISV000007

Parole chiave:

Forensic phonetics, Audio forensic, Noisy recordings, Speech perception, Psychoacoustics, Speaker comparison, Speaker recognition

Abstract

The paper describes a case study of particular interest and representative of speaker identification and speech attribution problems in real environments. The authors were recruited as technical consultants by the accused (already convicted) person’s lawyer to report on a probable misidentification of her client in two environmental recordings, with the goal of obtaining a scientific analysis and subsequently requesting an eventual review of the sentence. The results, carried out by the experts blind to the overall criminal proceeding, led to a conclusion of rare clarity in this field, and to the evident presence of a clear mistake in the speaker’s voice attribution. It is believed that the description of the work (albeit anonymized) may be of strong interest to those working for the promulgation of audio forensics and forensic phonetics scientific methodologies in real contexts such as preliminary investigations and legal proceedings.