With the help of special teams, the investigated voices, which are already acquired across interceptions of calls or common recordings. They are submitted to analyses that allow us to identify the characteristics of the voice, with the end to find a supposed culprit. Every day, the experts in "Forensic acoustics" take as a study object recordings of supposed extortionists, kidnappers, terrorists, etc. The analysis of voices that is realized in this area, (with judicial previous order) begins with the exact transcription of the dialogue contained in the tape. Some peculiarities like the elimination of phonemes, regional words, between others, must be born in mind by the forensic audio laboratory at the moment of the transcription of the tape. The transcriptions must be exact and the tape cleanliness in order to individualize the voice, because the morphologic characteristics of system of phonation are only in the human being. The dialogue is studied analyzed in different perspectives such as linguistically, semantically and grammatically , what verbs it uses and how it uses them, if it omits phonemes. With judicial previous order, the cassettes are submitted to studies to determine how there is the original sound of each of the vowels and consonants since they go out of the vocal cords. So, changing the accent, having flu, speaking slowly or more rapid is not inconvenient to identify a supposed culprit.