Conventional wavelet deconvolution ``flattens'' the signal frequency band and amplifies random noise. We present a method in this note that ``flattens'' the wavelet (signal) frequency band without amplifying random noise. It is a modified version of conventional wavelet decon. The method uses a decimated autocorrelation to compute the inverse wavelet at a coarser sampling interval. The order of decimation depends on the signal frequency range, relative to the full Nyquist frequency. After decimating, the algorithm fills with zeroes in the computed inverse wavelet to mimic the original sampling interval to obtain a new inverse wavelet. The inverse wavelet as computed has the desired inverse spectrum in the original signal frequency band. However, this zero-filled inverse wavelet carries a duplicate spectrum. Then we do a lowpass filtering on the zero-filled inverse wavelet to remove its duplicate spectrum, obtaining the final inverse wavelet. Convolving this final inverse wavelet with the original seismogram ``flattens'' the signal frequency band without amplifying random noise.