About the polyphony for Giga:
I don\'t know much about software for Giga and maybe not enough words to say here (who knows how the Giga\'s supposed to work correctly). However, as far as I know the voices has determined many variable factors. It\'s not only depends on the polyphony (sample mapping) when you play, it\'s also related to the way how you play it, and maybe how fast the tempo is?...
When you look at the GM-MIDI event (in cakewalk for example), you will see all kind of events like Note, Pan, Vol, Pitch wheel, sysEx, Expression, after touch, etc. Who can tells which elements will takes additional voices and which one is not compatible with Giga. Some of those events giga just ignores it but maybe some kind of illegal elenments, it can confuses then \"panic\" or just running up to the limit of 160 voices then might be processed to No. percentage of voices and defintely give up!
One more thing is we don\'t realy know how Nemesy comes up with 160 voices, what kind of the specification factors? Anyone knows? is it real 160 voices?
Well, I can not simply say it\'s wrong, since someone here can run it up to 160 voices without problems. But everyone run difference enviroment (Soundcard and PC). We need to come up with standard ways like: CPU type (PIII,Athlon) hardware system (500Mhz/256M/IDE-HD-7200), type of soundcard-Type (GSIF, none GSIF, multi-client, etc), GM-MIDI file (for testing), etc. (the guide line for forum and avoiding problems as well)
Just a thought! Maybe someone want to add more things here!