Hi All, anyone has used the implementation of PEAQ by McGill Univ.? I downloaded the latest version written in Dec. 2003, and modify the PQCB function according to its reference: "An examination and interpretation of ITU-R BS. 1387", since the original PQCB function is wrong. And then I employed the test senquence of acodsna.wav (at this moment, I have only this sequence and arefsna.wav), the final results as following:

Model Output Variables:
BandwidthRefB: 530.309
BandwidthTestB: 530.309
Total NMRB: 325.704
WinModDiff1B: 7.69139
ADBB: 0
EHSB: 0.233797
AvgModDiff1B: 4.59462
AvgModDiff2B: 50.2376
RmsNoiseLoudB: 0.0770393
MFPDB: 0.0146362
RelDistFramesB: 1
Objective Difference Grade: -3.907

It is greatly different to the results in Table 22 of BS. 1387, and who can tell me the reason? Do you meet this problem? Thank you very much.

First you posted at wrong place, this one should be in Scientific/R&D Discussion.
Second, why do you think that it is incorrectly implemented? Did you read P. Kabal's paper about PEAQ ITU-R BS.1387-1?

It's been moved.

youjy and anyone else reading who doesn't already know:
You are expected to read and understand the Terms of Service before posting.

Thanks a lot for your kind remind, I will read the Terms of Service immediately.
In addition, (to muaddib), could you please tell me the title of P. Kabal's paper about PEAQ ITU-R BS.1387-1, thank you very much.

From readme.txt from PQEvalAudio (McGill Univ. implementation of PEAQ):

Thank you. Maybe you mean this paper: "An examination and interpretation of ITU-R BS. 1387" P. Kabal, I have not carefully read it, and I just read the section of F.2 Critical Band Parameters, and change the original routine. The current implementation is written in Dec. 2003, and that paper and contained routines are written in Mar. 2002. So, I feel the current implementation should be correct.

Could you please give some advice, thank you very much.