How they should proceed if the resources are there
My thinking is that this would be a task similar to trying to crack 512-bit encryption using a consumer grade CPU.
In light of what I already wrote on this topic, if this company were able to develop the several other types of synthesis and incorporate them all into a mega synthesizer, and then use their genopatch technology to hard-crunch all the numbers for all the possibilities (too many to list, but you guys know what they are), how long would it take the best consumer grade CPU to come to an acceptable sound to the original as recorded in the sample?
Yes, it's feasible, but out of sight, presently.
But since it's not at all difficult to figure out the math on how to proceed, a solution is doable. Then, to shorten the amount of time doing the long way, crunching all the numbers the hard way, then smarter developers, and I suppose AI, figure out ways to crack the encryption using smarter shortcuts (Is the starting point a triangle wave, or a square? Thus, eliminate the all the others that are clearly incorrect, would shorten the number crunching. But if the first tone is a complex sample rather than a simple waveform....way more difficult, if not impossible).
Like my first post' sentence, pretty cool and fun. If you had the recipe from the Genopatch on how "that sound" was actually created, would give players an amazing tool to use in their everyday musical playing job.
One last comment - I think of my Kurzweil workstations and how complex and sophisticated the technology is, and nobody I know ever used a thimble of the machine's potential, trying to come up with the synthesizer that the genopatch must fundamentally have before it can ever get to work, is why I mention the astronomical uphill task ahead of this thinking.
Still, cool to contemplate.
EDIT: No, this is my last comment. lol The market for this "SUPER SOPHISTICATED" synthesizer would have almost no sales given how ridiculously difficult it is to program. Most people don't tackle programming on Kurzweils for this very reason. In any case, the market for this product would not be for the purpose of selling a super synthesizer to program from scratch (though there are those who would), not nearly as much as the need for the super synthesizer to be in place in order to use the GenoPatch and bring all the sophistication to a useful purpose - the end result for users. That would be the result worth R&D and for the end user, worth buying.
EDIT 2: Eliminating the need to allow users to program this super synthesizer would make this extremely complex software synthesizer much easier to develop, by comparison. The only reason for it is to recreate sounds fed to it from samples, giving them the recipe it came up with. But if you need to allow users to have access to controls to program it themselves, that's where the nightmare for the Company's engineers begins. So theoretically, 2 versions. One for recreating, at one price tag. And the other, which is WAY more complicated for the engineers, for users to also program. at a much higher price tag.
Last edited by Toddskins; 22nd January 2024 at 12:45 AM..