convert to signal and write into buffer?

Yes that’s the direction I have been going in, but what I'm really trying to work out is how I can get max to analyze some audio from different samples and then categorize it based on similarities or differences in pitch slop and intensity. E.g. if a sample has similar data coming from its pitch it will be named x or y or not recognized. But i suck with max so cant get it working.

this might not be answering your question at all, but then it might be useful nonetheless: http://www.e--j.com/?page_id=499 is a package that can describe sounds in terms of fairly abstract concepts such as spectral slope etc

if you can coerce the output of these descriptors into a file or some other data structure (ie coll) you may then be able to compare them...