The Wavelet convolution attribute allows you to convolve a wavelet with the select seismic data, where the wavelet can be considered as or act as a filter.
Convolution is a mathematical operation on two functions that is the most general representation of the process of linear (invariant) filtering. It can be applied to any two functions of time or space (or other variables) to yield a third function, the output of the convolution. Although the mathematical definition is symmetric with respect to the two input functions, it is common in signal processing to say that one of the functions is a filter acting on the other function.
The convolution and the following operations are preformed in Frequency domain.
Normalization : Application of the Parseval’s theorem before convolution.
Scaling: Multiplies the amplitude of the frequency components by the selected value. (default: 1)