Speech analysis has a type called the voice stress analysis that detects stress in the form of micro tremors in the speech of a person. It is important to note that the voice stress analysis is equivalent to polygraph. Both polygraph and stress have a common problem of not being able to point out deception. This notwithstanding, the speech analysis is apparently being used by both the Federal Bureau of Investigations and the Central Intelligence Agency of the United States. The fact that each and every person possesses a unique voiceprint makes the identification of a speaker to be theoretically possible. This can be achieved through acoustic analysis of the intonation counters, pitch and vowel length. When the pairs of the spectrographs that show the suspect speaker and the known speaker uttering similar words are matched, the speaker can be easily identified (Olsson 2004, p. 71).
The fact that voice stress analysis relies on eye comparison is a big problem. Another problem involves the variation that occurs in the same speaker. It is reported that the uttering of the same sentence a hundred times in quick succession does not produce any two identical uttering. Some countries like the United Kingdom, however, prefer auditory analysis as opposed to the acoustic method. In auditory analysis, the speech samples are phonetically transcribed. This analysis is important as it allows the analysts to identify such features that are idiosyncratic like the speech impediments and the unusual realization of phonemes. Besides, the analysts might find the need to profile the social and regional identity of the speaker. Speech analysis nowadays accepts the mixed method as the most accurate and reliable. It can found its application in situations where the suspect or the accused possesses both the elongation and a block stammer, which are associated with fricatives and stops respectively.
Another important type of voice analysis is the voice line ups. Here the suspects that have been picked by an eye witness are used. A case in point was during the 1933 Lindberg case in which a boy was kidnapped and killed. The father to the victim had apparently spoken with the kidnapper concerning the issues of the ransom and therefore, claimed to be able to identify the voice. However, it is reported that the frequency by which listeners are able to report familiar voices, can be doubled by the frequency of the mistakes in reporting such tones. Moreover, the ability of an individual to recognize voices is hugely varied. According to an experiment aimed at testing the famous voices by some participants, the success rates varied hugely from forty six per cent to a hundred percent (Coulthard & Johnson 2007, p. 241).
On the other hand, the ability to effectively recognize a voice can be hugely affected, in a detrimental manner, by a delay. A success rate of eighty seven per cent within two days has a potential of decreasing to thirteen per cent after five months....