On the utilization of overshoot effects in low-delay audio coding

In the speech coding community a low-delay coding usually means that the coding delay should be less than 5 ms, e.g, 2 ms in ITU G.728 [2]. It is reasonable to adopt this rule also to low-delay wideband audio coding. An acoustic signal propagates in air less than one meter within 2 ms and therefore this low coding delay should not produce significant audible echo problems for example in high quality teleconferencing or in applications based on shared On the other hand, 2 ms is 88 samples of an audio signal at a detailed auditory spectral model for an onset of a musical signal coding scheme evidently fails in the cornpared to a steady-state situation where it is possible to use past In low-delay audio coding (coding delay < 5 ms) there is no time for detailed spectral modeling in the case of brief percussive sounds, e.g., the castanets, and onsets of music Or speech sounds. On the other hand, it is known from psychoacoustic experiments that the ear is not accurate near the onset of a wideband sound. In this paper, we study the audibility of coding errors near the onsets of musical sounds in a simulated low-delay audio codec based on musical transients it is sufficient to reproduce a rough temporal ms. Preliminary listening tests support this idea. It is proposed that the overshoot effect of hearing could be utilized efficiently in acoustics. frequency-warped linear prediction. It is suggested that for many kHz sampling rate and it does not make to estimate and spectra1 envelope Of the Original during the 5-10 within this interval. Accordingly, the performance of a low-delay of a sudden enhancing the performance of a low-delay audio coding scheme. signal ,,dues, Nevertheless, it is known from several different lis-