3 ms·
what if you take the windowed cosine transform, and then see where the transform is within 10% of each other (or less than a threshold value). Then zero out al
by dnautics 13y ago
what if you take the windowed cosine transform, and then see where the transform is within 10% of each other (or less than a threshold value). Then zero out all such regions. With the technique you're using I'd worry about the phase happening to be exactly 180 degrees if the frequency and position of the noise source are 'just so' - which might explain your observation of 'pops'.