TTU etherium

[Touch The Universe]

I get your Point my friend, but I think you are missing mine. Is not sampling per se, as mostly samples are just set the dials then sample it, but I want to sample every single parameter of a dial and then have a knob interpolate between those samples with every single parameter knob turn. Now let me know if this is what you were talking about with extra dimensional weight tables, cuz I do not know

[Touch The Universe]

The only possible way for this to work, is not with a very small weight table, but it requires at least one or two seconds per each value in a knob, so capture ing all it's perimeter, so whatever you want to call this thing, one knob could easily equal 100 MB or maybe even a gigabyte oh, this is where it makes sense to be smart, utilizing compression, it might even make sense to lose a lot of quality, and use flac or maybe even MP3s, especially if I want to implement cross perimeter, ie capturing a knob at 10, with a knob at 1, so essentially we are multiplying 128 by 128 so we are capturing to knob perimeters.one word describes this, it is quite ludicrous, quite crazy but we are living in a world where there are 1 terabyte sample libraries, I think we are at a stage where we should start embracing crazy ideas like this because it is technologically feasible

[CableChannel]

I'm not fully clear about the concept either. Maybe you could name more specific parameters that you want to sample as a wavetable. I guess you are not talking about filter / resonance / attack / release and that sort of stuff, right? I don't know what the formant synthesis knob that you mentioned does. Is it a filter?
Wavetables are normally used to store one cycle of a waveform, so I guess one can't store the resulting sound of a subtractive engine in it. And even if it were possible, with parameters like cutoff, resonance etc., you would need to sample all 128 cutoff settings for each of the 128 resonance settings, and all of that a 128 times for a third parameter, and so forth, making up the multidimensional parameter space that people mentioned here. Even on nowadays computers, this concept would quickly break all barriers of available realtime processing power.

[dmbaer]

Regarding multidimensional wavetables. Who knows, but there are already synths having 3D wavetables (standard wavetable with at least 1 additional dimension like Icarus). And knowing you already can blend 2 or more wavetable oscs, the magic is gone.

I don't have Icarus (either 1 or 2), but I did just download the manual for Icarus 2. It's a most impressive instrument, but its wavetables are *not* 3D. Conventional WTs are a one-dimensional array of waveforms. Continuous morphing involves averaging the frequency domain characteristics of two adjacent waveforms, weighted by the index position (the closer you are to one, the more influence it has in the resultant waveform).

A 2D wavetable would be a two-dimensional array of waveforms and morphing would involve four waveforms based upon two orthagonal indicies. 3D would add another dimension and morphing would involve eight waveforms and three indicies. We don't talk about 4D because we don't want anybody's brain to explode.

Icarus 2 wavetables are ordinary one-dimension arrays of waveforms. That's not a 3D wavetable.