Hi Dan, Hi list! I was thinking about a little ChucK convolution patch, kind of as a excersise, kind of as a joke. Basically I was thinking about taking a array of SndBuf's (as many as there are samples in the impulse responce and all holding the same impulse responce) and firering them off in turn each samp with a gain determined the .last() of the input sound. Simple, fun and probably a bad idea on any practical level since excelent free convolution reverbs are around already. Then I thought of LiSa... LiSa is already a lot like a array of identical buf's so this might even make a bit of sense. Sadly it turns out that LiSa lacks something like this; mylisa.voicegain( voice, float) The more I think about that, the more it makes sense as a practical addition, not just for convolution. For example this might be mapped to velocity for some cheap & cheerfull soft-sampler and per-grain variable volume is usefull in grain-clouds as well. Could this perhaps be considdered for a future update? Cheers, Kas.