Masking models that evaluate the audibility of error signals have a limited validity for assessing perceptual quality of parametric codecs. We propose a model that transforms the audio signal into an Internal Representation (IR) consisting of temporal-envelope modulation patterns. Subsequently, the IR of original and encoded signals are compared between both signals. Even though the audio signals compared may be uncorrelated, leading to a large error signal, they may exhibit a very similar IR and hence are predicted to sound very similar. Additional post-processing stages modeling higher-level auditory perceptual phenomena such as Comodulation Masking Release are included. Predictions are compared against subjective quality assessment results obtained with encoding methods ranging from parametric processing methods up to classic waveform preserving codecs.

Authors: van de Par, Steven; Disch, Sascha; Niedermeier, Andreas; Burdiel Pérez, Elena; Edler, Bernd
Affiliations: University of Oldenburg, Oldenburg, Germany; Fraunhofer Institute for Integrated Circuits IIS, Erlangen, Germany; Friedrich Alexander University, Erlangen-Nürnberg, Germany; Fraunhofer HAS, Oldenburg, Germany(See document for exact affiliation information.)
AES Convention: 147 (October 2019) Paper Number: 10314
Publication Date: October 8, 2019 Import into BibTeX
Subject: Product Development
Permalink: https://www.aes.org/e-lib/browse.cfm?elib=20686

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.

Learn more about the AES E-Library

E-Library Location:   /conv/147/10314.pdf

Start a discussion about this paper!