by Carson Reynolds
So for a while I’ve been kicking around the idea of using wavelets to decompose a track into different resolution components. I want to take the low resolution components of one track and layer a higher resolution track on top. This should be somewhat akin to pulling out the deeper atmospheric components of a track and then tossing the finer nuances of another on top. Well today I finally got around to playing a bit. Taking a source track, I computed a really low resolution version, and then a track with an approximation and detail decompositions after one another.
Seems promising. I’ll try to put a whole track together this week using this technique. An interesting permutation on this might be to use the details computed for one track’s approximation to regenerate the signal using a second track’s approximation.