Adaptive transform coding and adaptive subband coding are widely used
methods for high-quality audio coding. Both coding schemes primarily u
se masking effects of the human auditory system for data-rate reductio
n. A further reduction can be achieved by coding noisy and tonal signa
l components separately. Methods to distinguish between tonal and nois
y audio components in the time domain, the frequency domain, and the s
ubbands are described. Combinations of these coding schemes with adapt
ive transform coding and adaptive subband coding are presented. A high
ly efficient implementation of adaptive subband coding using noise det
ection in subbands is proposed, and results of this coding scheme are
given.