Hi Italo,
thanx for the swift reply. You're probably right about different code running differently on similar DSP, depending of course on lots of factors. In the Clavia modular synth concept, every little module you use when you design a patch eats some DSP processes, some i.e. a two-input mixer eat very little but a vocoder eats a lot (it uses something like 40% of the available DSP resources). So, you can't really do a lot, if you use two vocoders. You might argue "who needs two vocoders" but well… you know… In the Clavia, if you add an expansion card which btw has another four DSPs, you don't really double the DSP power, you just have twice the polyphony. Since the Eventide offers so much more in terms of modules you can easily go wild: some of the stuff in other open ended architectures, like CSound, can be very DSP consuming (physical modelling comes to my mind here as well as granular synthesis). Of course you can always try optimization techniques, this is always very good as it unclutters the patches thus making them run more efficiently, but how far can you go without cutting corners?
Some ages ago I requested a demo CD from Symbolic Sound and they had an amazing effect there… it was a voice that was being recorded in the sample memory, which was limited to maybe a few seconds but that was on purpose. The algorithm granularized in real-time the audio as it was being continously recorded into that memory. It was a dry-wet balance with some of the audio coming directly from the voice whereas the wet signal was the tiny granules sprayed all of the over the audio field. The usual suspects, like panning and reverb were there as well. These granules were like they were floating here and there.. Lol, I may be over-poetic here but it was an awesome effect! I am pretty sure it can be done in the H- 8000 FW, Italo tell me what you think. It will certainly be the first FX I will design in Vsig!
Regards
Yannis
