Sidelobe levels of weighting functions with a platform when the number of samples (N) is small
F.E. Churchill · Proceedings of the IEEE · 1984
The highest sidelobe level is a figure of merit of weighting functions. It has been found that published figures [1] for weighting functions that rest on a platform, e.g., Hamming, Blackman-- Harris, apply only when N is large. The sidelobe levels are significantly higher than the published figures when N is small.