A software to evaluate the quality of noise-corrupted audio signals is described. This software can be used to assess the perceived quality of coded audio signals, in effect assessing the quality of the coder. Given an original signal and a noise (coded) one. PERCEVAL will compute the probability of detection of the noise as a function of time. This software is intended both as a development tool for coder design and as a means of assessing the perceived quality of coders. The subjective quality criterion, which is the heart of this software, is based on an auditory model developed by the authors' group. This auditory model has been tailored and tested independently of PERCEVAL on simulations of psychoacoustical experiments described in the literature. This model has a fast algorithm, which could result in a real-time implementation on today's digital signal processors. Both the software PERCEVAL and the auditory model it is based on are described. Test results are included for both.
Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!
This paper costs $33 for non-members and is free for AES members and E-Library subscribers.