This paper addresses the problem of equalizing an audio signal in a constantly changing noisy environment. The purpose of equalization is to provide perceptually equal loudness of sound regardless of the environmental conditions. Based on an automatic estimation of noise level and its spectral content, selective amplification of frequencies masked by noise is performed. In the case of speech signals, the result is intelligible speech regardless of the surrounding noise. For musical signals, an improved comprehension of the musical content is achieved.
This paper costs $33 for non-members and is free for AES members and E-Library subscribers.