In this paper, a real-time implementation of the loudness model of Glasbergand Moore [J. Audio Eng. Soc. 50, 331‚Äì342 (2002)] for time-varying sounds is presented. This real-time implementation embodies several approximations to the model that are necessary to reduce computational costs, both in the time and frequency domains. A quantitative analysis is given that shows the effect of parametric time and frequency domain approximations by comparison to the loudness predictions of the original model. Using real-world music, both the errors introduced as a function of the optimization parameters and the corresponding reduction in computational costs are quantified. Thus, this work provides an informed, contextual approach to approximation of the loudness model for practical use.
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.