Greetings all,

In my recent work I've been working with the wonderful PeRColate library. Unfortunately, I had discovered that the Pd/Linux version of this library has a bug in the cool "munger" object that is capable of real-time granular synthesis. By experimenting with it, it seems that the parameter values are not interpreted linearly, but rather, what would be a range of 1-100 is on linux something like 0.0000001 to something like 0.0001. Without looking at the source code, I am not sure whether the values are distributed linearly and just the decimal point is out of whack or if the values are also distributed in some weird fashion. The best way to test this is to take the example patch and try to apply some of the values that are found as presets in the MaxMSP version and it will become quickly obvious what I am talking about. For this reason, munger also quickly eats up all the CPU as the number of grains parameter tends to encumber the processor for obvious reasons (i.e. value of 1 sent to this parameter in Linux version could be equivalent to something like one million grains or something similarly preposterous).

I am very much interested in using munger in conjunction with Pd and for this reason I was wondering whether anyone on this list was familiar with the port of this lib and was possibly able to look into this bug. I'd be happy to assist as much as my time allows which unfortunately until the end of this calendar year won't be much at all :-(.

Best wishes,

Ivica Ico Bukvic, composer & multimedia sculptor http://meowing.ccm.uc.edu/~ico/