nikolajandersson / peaq

This repo contains a Matlab version of the Perceptual Evaluation of Audio Quality algorithm

PEAQ

This repo contains a Matlab version of the Perceptual Evaluation of Audio Quality algorithm.

The idea was to isolate the algorithm and update the code so it can run in newer versions of Matlab. I am not maintaining the code and the code is provided as it is!

The algorithm was originally implemented by Peter Kabal in 2002 and the Matlab code is fetched from jorgehatcrrma's repo.

Resources:

Kabal, Peter. "An examination and interpretation of ITU-R BS. 1387: Perceptual evaluation of audio quality." TSP Lab Technical Report, Dept. Electrical & Computer Engineering, McGill University (2002): 1-89.

Wiki

PEAQ – AN OBJECTIVE METHOD TO ASSESS THE PERCEPTUAL QUALITY OF AUDIO COMPRESSED FILES

PEAQ -The ITU Standard for Objective Measurement of Perceived Audio Quality

Python version by Stephen Welch and Matthew Cohen

peaq's Issues

How to run for .wav with Sampling frequencies different?

Can I run the code for samples having sampling frequency other than 48 kHz supported?
Can I run the code for comparing samples having two different frequencies?

What is PQFB?

get confused by PQFB in line 16 of PQadapt.m and line 17 of PQloud.m, seems like a function like PQCB() but not implentment

Reference

I think the algorithm actually comes from this paper by P. Kabal in 2002. It would be nice to have this cite on the readme:
Kabal, Peter. "An examination and interpretation of ITU-R BS. 1387: Perceptual evaluation of audio quality." TSP Lab Technical Report, Dept. Electrical & Computer Engineering, McGill University (2002): 1-89.

Note that this is mainly about the last name being 'Kabal' instead of 'Kabel'.

odg value negative

Hi!
Thanks for the repo. Much appreciated.
I was testing the code for two audio files and the odg value comes out to be something negative. As I understand, PEAQ values range is from 1-5 to mimic the human user. Can you verify if something is wrong with the code? Thanks!

Add sample wavs to repo -- so that it is seldf contained.

Ideal pair: diode clipper with 16xFs anti-aliasing and 4xFs with or without anti-aliasing.
Thanks -- Cumhur