No abstract

Read and spontaneous speech was produced by 100 male adult speakers of German in a neutral setting and in a Lombard setting, where 80 dB noise was presented over headphones. Average f0 (‘f0mean’) and relative standard deviation of f0 (‘f0varco’) were determined for each speaker in each of the four conditions spontaneous neutral, spontaneous loud, read neutral and read loud. The results confirm that when increasing vocal effort from neutral to loud speech, f0mean increases as well. None of the 100 speakers posed an exception to this effect, but the size of the effect differed between speakers, even after differences in amplitude level were accounted for. F0varco was significantly higher in loud than normal speech for the reading task, whereas in spontaneous speech no significant difference occurred. These results are compared with the literature and discussed with respect to explanations and forensicphonetic implications.

. InABSTRACT Expert witnesses in phonetics find themselves more and more often in forensic situations in which they have to identify the voice of a speaker who does not speak their native language. Until recently, little has been known about the role which the nativelanguage background of the listener plays in such speaker identification tasks. In this report, several aspects of an experimental investigation on the influence of native-language background on speaker identification are reviewed. Results of a first experiment are reported and some follow-up experiments currently being carried out are described within that context.

One of the most significant features in forensic phonetics used to identify a speaker is his or her so-called regional accent. In a speaker profiling task, German experts for voice comparison were evaluated concerning their performance and their methodological approach in accent identification. As a result, the experts as a group performed well; only very few severe confusions of regional accents occurred. Results also showed that the experts applied different strategies in identifying a speaker´s regional accent; the different methods were (a) aural-perceptual analysis, (b) acoustic analysis, (c) consultation of audio databases, and (d) consultation of literature on dialects. Participants also differed considerably in the amount of time they needed to complete the task. The most important finding is that no correlations existed between the performance (i.e. error rate) of the participants and (a) the time spent, (b) the number of methods applied, (c) the kind of method applied or (d) the perceived degree of difficulty. This leads to the conclusion that the performance in dialect identification may predominantly depend on the individual (innate or trained) skills of the expert in speaker identification.

In various applications of forensic phonetics the question arises as to how far aural-perceptual speaker recognition performance is reliable. Therefore, it is necessary to examine the relationship between speaker recognition results and human perception/production abilities like musicality or speech sensitivity. In this study, performance in a speaker recognition experiment and a speech sensitivity test are correlated. The results show a moderately significant positive correlation between the two tasks. Generally, performance in the speaker recognition task was better than in the speech sensitivity test. Professionals in speech and singing yielded a more homogeneous correlation than non-experts. Training in speech as well as choir-singing seems to have a positive effect on performance in speaker recognition. It may be concluded, firstly, that in cases where the reliability of voice line-up results or the credibility of a testimony have to be considered, the speech sensitivity test could be a useful indicator. Secondly, the speech sensitivity test might be integrated into the canon of possible procedures for the accreditation of forensic phoneticians. Both tests may also be used in combination.