it's also not a very clever one. I think what the author wants to do is actually segmentation which is much more complex program. the best algorithm Ive used fo

it's also not a very clever one. I think what the author wants to do is actually segmentation which is much more complex program. the best algorithm Ive used for doing any kind of amplitude following is a detrender with two envelopes running at different rates. next to that is Foote's audio novelty audio algorithm.