Sorry for the long delay in getting back to this. Adam Schabtach writes: >>OTOH, Carlos herself abandoned analog synths and switched to digital machines to do her most accurate imitative synthesis work, "Digital Moonscapes", and her most exploratory, "Beauty in the Beast".<< Then Carlos abandoned digital and went with samples. Carlos' additive synth percussive sounds are really good, but most of the pitched instruments aren't very good, that's because there is no fixed formant filter bank on the Crumar (the very thing we're talking about). Compare my Nord violin with Wendy's and there is no comparison. We basically haven't really heard what a synth can do which is part of my point. >>Frankly I find it kind of amazing how much effort people will put into talking about what other people do with their synths. I mean, really, wouldn't you rather spend the time playing your synth?<< I play my synths. But when somebody mentioned an advanced fixed filter bank and how much of a market there would be for it, that involves how people use their synths. If they're not up to doing sophisticated filtering, then a product like that isn't going to fly. Mark writes: >>I'm more for practical modules than esoteric modules or modules that do things that are readily available elsewhere.<< And thus the lack of esoteric modules, which then leads to less interesting synth sounds. Practical modules already exist everywhere. There isn't a single advanced filter bank for synthesis unless you get into using Soft synths on a computer, and then you lose the immediacy of the modular synth interface. >>Imho, imitative synthesis requires imitative playing. Even if you could create a violin patch that was indistinguishable from the sound of a real violin, you would need an awfully sophisticated controller to get it to sound anything like a person playing a violin. I severely doubt anyone could accomplish it with a CV/gate sequencer without a massive amount of steps and channels. So if you are already using MIDI to control it, and you are using an editor/librarian to edit the parameters on this proposed module, then why not just do the whole thing in a computer??<< It seems like so much of what I had said in my last post wasn't remembered. I had mentioned that I apparently got enough expression out of my Nord violin playing in real-time (just using velocity and a pedal) to make some people doubt I was playing it in real-time and that I must be using a midi sequencer somehow (or playing a real violin through the Nord), and I still have a hand, another foot my mouth free for even more expression. The Garriton Stratavari violin sounds like the real thing in real-time just using keyboard and modwheel. That's what I'm trying to accomplish. Acoustic musicians use similar phrasing and expression in many different instruments. The Yamaha physical modeling synth uses a keyboard/breath controller or a wind controller for all of it's sounds. That kind of control works just as well for brass as for woodwind, so it would work just as well for the imaginary new synthesized acoustic instrument. >>That's my point, if something is DSP, then it should have reasons to justify being in a separate piece of hardware -- interface, stability, copy-protection, portability, etc.<< Interface, patchability and elimination of processing delay and of computer. Real knobs, rather than mouse and screen. Some of the same reasons people use real modulars over soft synths in the first place. >>While I agree that there are people who buy synths, and perhaps even modulars, who use them in limited ways, I doubt that most people who buy modulars don't produce music. How well someone can play an instrument doesn't have much to do with it.<< Judging by what I've heard, many if not must don't use them to produce music. Many aren't musicians which is one reason why. And you're contradicting yourself. How well someone can play an instrument DOES have to do with it. If somebody can't play a synth like a musical instrument, they probably aren't going to be interested in a product that's primarily for synthesizing complex musical instrument type sounds. >>By all accounts, Verdi could barely sing and Shakespeare was a lousy actor.<< And thus Verdi didn't sing and Shakespeare didn't act, they got other people to do that. Unfortunately those who can barely play an instrument or compose music don't follow these same rules. >>Most of of today's electronic music uses sequencers -- either hardware or data in a DAW.<< And how does that music get into those sequencers? Usually it has to be played into them in the first place, then editting can be done where needed. I'm well aware of the flood of simple, robotic and repetitive stuff out there. An advanced filter bank isn't for the that latter group, since a melody or chordal structure needs to be there for filtering in the first place. >>Regardless, I find your argument -- that you found all kinds of bugs in the latest synths you bought is evidence of how people use modulars -- rather non-sequitur.<< It's completely relevant. Simplistic or limited use of the features and filters that currently exist in a modular means people aren't in the market for something that's way beyond what's already there. Just the arguing people are doing with me that an advanced filter bank isn't needed proves my point. And that's why there is no such thing and probably never will be. So the few of us who want to experiment in that area have to use less than optimal alternatives. -Elhardt
Message
Re: [motm] Re: Imitative Synthesis and Implications for Hardware
2007-05-13 by Kenneth Elhardt
Attachments
- No local attachments were found for this message.