You just store the peak amplitudes (+ve and -ve sepratly) for every N samples.
I think typically you have two seperate peak files, eg. one for every 2^8 samples and one for every 2^16. The eqivalent in large images uses one "peakfile" for every power of two, but audio people seem to be more conservative - for one thing scaling audio is much cheaper. You could look at sweep, snd, ardour, audacity etc. to see how it is done. - Steve On Mon, Oct 14, 2002 at 10:45:20AM +0200, Thomas Grill wrote: > Hi all, > can anyone give me pointers on how the overview cache for a zoomable > waveform display is organized? > One can see accurate and fast displays in a lot of applications but i guess > the rendering of this is not straightforward. > > best greetings, > Thomas
