Document Fragment View
Fragment Information
Showing contexts for: spectrography in Cbi vs . Anup Kumar Srivastava Etc. Judgement ... on 26 August, 2016Matching Fragments
70. A question has been raised about the possibility of error in identification/ comparison by the expert. This witness was thoroughly crossexamined and I would like to reproduce relevant portion of crossexamination by Sh. V. K. Ohri, Ld. Defence counsel for A2 as under :
"Question : Is it correct that there is an error in the range of 0 to 80% in auditory examination of voice for comparison?
Answer : Depending upon the quality of the voice, that is, if a person mimics or disguises his voice while giving sample, the percentage of error may be more, that is, upto 80%. But, if the sample quality is good, that is, given in the manner a person speaks, the error would be very less and could be even 0 error. The error in auditory examination is further verified by objective method, that is, voice spectrography, which has accuracy and reliability of voice comparison to the extent of more than 99% even if a person mimics or disguises in his sample voice.
73. Ld. Defence counsel has assailed his opinion on the ground that there could be error upto 80% in such opinions. Ld. Defence counsel has referred to following portion of the crossexamination, which is reproduced as under :
"Question : Is it correct that there is an error in the range of 0 to 80% in auditory examination of voice for comparison? Answer : Depending upon the quality of the voice, that is, if a person mimics or disguises his voice while giving sample, the percentage of error may be more, that is, upto 80%. But, if the CBI Vs. Anup Kumar Srivastava etc. Judgement Dated : 26.08.2016 sample quality is good, that is, given in the manner a person speaks, the error would be very less and could be even 0 error. The error in auditory examination is further verified by objective method, that is, voice spectrography, which has accuracy and reliability of voice comparison to the extent of more than 99% even if a person mimics or disguises in his sample voice."
74. From this answer, two facts can be understood. First, there are chances of error. Second, with the help of a scientific instrument i.e. voice spectrograph, more than 99% accuracy is achieved in identification of voice. In the present case PW16 has used voice spectrography and has given a very reasoned report, wherein he has not given opinion on certain portions of the questioned voices.
75. In view of this discussion, I hold that the voice identification report of PW16 is scientific in nature and has to be accepted.