Comments on ''On the computation and the effectiveness of discrete sine transform'' (Q1082057)

These comments are concerned with the paper by \textit{P. Yip} and \textit{K. R. Rao} [ibid. 7, 45-55 (1980; Zbl 0456.65079)]. The effectiveness of the discrete sine transform (DST) in terms of residual correlation is re- evaluated. It is concluded that for a Markov-1 signal with low correlation coefficient \((| \rho | <0.6)\), the version 1 of the DST (DST1) is more effective than the discrete cosine transform (DCT), and that for high correlated Markov-1 signal with negative coefficient \((-1<\rho <-0.85)\), the version 2 of the DST (DST2) is superior.