Comment on "Parseval relationship of nonuniform samples of one- and two-dimensional signals"

For original article see Marvasti and Liu (IEEE Trans. Acoust. Speech, Signal Processing, vol.38, no.6, p.1061-3 (1993). In the original article the proof of (1) states that x/sub lp/(t) is the low-pass filtered version of the nonuniform samples and can be obtained by the following interpolation: x/sub lp/(t)=/spl int//sub -/spl infin///sup /spl infin//x(t/sub n/)sinc[W(t-t/sub n/)]dt. The present authors suggest that this equation is incorrect, because the response of a linear system to an arbitrary input signal is the convolution of the input signal and the impulse response df the system according to the signal and system theory. Thus, they give a revised version of x/sub lp/(t).<>