Further to (but different from) my previous post on Vocal detection, I was wondering whether the sections data (or a new set of data) could provide some information on stereo variance in a track - i.e. simply compare the two channels and calculate a variance level (with 100% being total variance and 0% being equivalent to mono).
Additionally this variance data might also/only (as an average could be taken) be split between ranges (ideally roughly equivalent to speaker ranges: 0-1k, 1k-2k, 2-20kh, 20kh+ (inaudible))
Additionally (although I'm sure you've covered this) but I haven't spotted a getChannels method or equivalent for mono stereo or....
Chrisbod - thanks for the suggestions about adding stereo variance in the track and for supporting stereo data in the analysis output. We'll add it to the list of interesting things to do.