AP Synthesis for user-PCM oscillators?

[Arghblarg]

Here in town (Calgary) Roland had a vendor's preview of their new gear, about 2 months ago.. it was meant primarily for music sales reps, but I got an email invite from Roland Canada, so myself & a bandmate showed up

They had a V-Synth GT beta unit, and the Roland reps were very eager to show it off. The guy demoing it showed me that one can swap in a different waveform and define whichever AP algorithm you wanted to apply to it. Granted he swapped one preset wave for another, but that sounds like AP algorithms can be applied to any wave.

Error 144 - Reboot y/n?

[s.asad]

^Yes, it details in the manual applying AP-Synthesis modeling types to other waveforms (square, sine, etc.)

I assume Roland will soon (perhaps they already have) include most types of encoded waveforms to utilize AP-Synthesis.

That is when I will buy a V-Synth GT. No sooner.

Maximizing the utility of a new flagship feature is essential to the marketability of said feature.

[Krisser]

I spent a little time with a GT on demo here in Belgium, and it seems pretty clear that AP can only be used with a set number of built-in waveforms.

On the plus side, this means that you are free to use a Sine with the Violin phrasing model to get unusual sounds, but on the minus side you cannot run any sampled or imported waves through AP. I think there are about 20 to 30 waves that can be mixed and matched with the AP articulation models, and how many unique sounds can be made with the available number of combinations remains to be seen.

I was thinking about the possible reasons for not allowing user waves to go through the AP synthesis. This is sheer speculation, but I wonder if it might be because the waves used for AP are actually several layers deep. So with the violin for example, there is a pizzicato layer, a bow scraping layer and so on. And what the AP does is access different layers of the sample depending on how the keyboard is played, as well as adding legato and so forth.

If this is the basic mechanism, then it can be seen that simply sampling your own flute would not provide enough info for the AP to do it's work.

Of course, I could be (and probably am) completely wrong about all this...
K

[s.asad]

^That sounds very possible.

But what I read(maybe I'm mistaken), indicates AP-Synthesis modeling types can be applied to all the stock waveforms.

That would necessitate each source waveform (including square, saw, etc.) having seperate layers for pizzicato, bow scraping, etc.

It is possible that is the case.
If so, that's not necessarily a negative thing.
Nor something that would dismerit the "revolutionary" tag given to AP-Synthesis.

A remedy for special AP-source-waveform-dependence, and a truly great edit feature, would be AP Layer encoding for variphrase encoded PCM-oscillators. Basically allowing for preset-processing env/algorithms to alter one encoded user wave several ways to suit each necessary layer for thorough function in AP-Synthesis.

I tend to assume that the modeling types are a seperate process/code, totally unreliant on a limited set of dependent waveforms.

Either way, I am very eager to see Roland make AP-Synthesis a truly revolutionary flagship feature for 2 reasons:

#1. So V-Synth sales increase, and lots of thaught, time and money can be invested in the next monster Variphrase flagship
#2. So I can buy one

[s.asad]

Has anyone determined whether or not AP-Synthesis is dependent on layer switching between specialized waveforms, or if it is actually a genuine organic synthesis method, literally modeling the waveforms upon play, instead of playing modeled waveforms?