ABSTRACT
In a number of criminal offences such as kidnapping, extortion, intimidation and nuisance calls the human voice or speech transmitted via telephone plays a devisive role. The aim of voice comparison is to show whether or not two samples pertain to the same individual. Strictly speaking, only statistical statements may be made in forensic voice comparison. In forensic voice comparisons, however, speakers normally are not co-operative, and may even disguise their voices. An auditory analysis by the trained ear of a speech scientist should not be dispensed with since its results wil not be obtained by any computerised techniques in the foreseeable future. In order to establish a statistical base for a reliable interpretation of the data large experiments will have to be performed, investigating the behavior of speaker-specific features under various conditions such as speaking situation, voice disguise, and background noise.
