Taika-Kim in this thread: http://www.kvr-vst.com/forum/viewtopic.php?t=15319 wants new synthesis and is waiting for the next revolutionary synth.

So far I don't think there have been any revolutionary designs, simply different ways at modeling sound based on the rules of physics. I also think there are so many new ways of synthesizing (spectral, waveshaping, granular) that the point is moot.

What do you think? Is synthesis built by revolutionary or evolutionary means?

X_Bruce quoth So far I don't think there have been any revolutionary designs,

What, at all? Ummm, no...

simply different ways at modeling sound based on the rules of physics.

Subtractive synthesis doesnt have much to do with physics. Neither does waveshaping.

Basically, anhtjing which generates a patter can be used as a syntheis methid. In academia it doesnt matter as much, but in 'commercial' areas its whether 'control' maps well to m that pattern thats important.

I also think there are so many new ways of synthesizing (spectral, waveshaping, granular) that the point is moot.

None of these are that new. The newest are probably about 20 years young.

What do you think? Is synthesis built by revolutionary or evolutionary means?

Both. Some 'revolutions' create the starting point of evolutionary development. Five years ago if I'd said 'granular synthesis' to you, you couldnt have found other references to it outside of academic research papers. And it was an old idea then. Now, its a 'big new thing' But its evolved a lot in that time.

Hi Guys,

V. interesting topic - with some really good points... but by definition (and I may be naieve here) isn't "waiting for the next big thing" actuall waiting for the next evolution of existing processes?

I mean sound, by definition and whichever way it is synthesized, is (just ) waves so is it not just the (next) evolution of the creating process?

If it were possible to say, create sound from from light (?) that would be a pretty wild way of synthesizing sound - but because sound and light already exist its the (natural) evolution of the process?

Or perhaps I'm totally wrong - Didn't get much sleep lastnite (sick) so I'm in work knackered

Cheers,

well, creating sound from light isnt anything new at all actually. infact its quite old now. anyone who messes with graphics might know that light in computers is actually RGB (CYMK being darkness - print). From this light we are able to create images, ie. bmps. from bmps we are able to produce sound, ie. Fruityloops Beepmap.

actually, Aphex twin made a sound on one of his albums using this kind of process. If you were to load the track up in say cool edit and view it in its spectral form you can actually see a face, which i think is his. but listening to it would just result in a weirdish kinda sound. which pretty much sums up his whole discography

Rock on!

i quite like the idea of the neural network synth. i think having this in vsti form would be kinda cool. from what i gather, your able to load in a sound, say from an album you like, zoom in on a specific sound, sample it and it is able to reproduce this sound throughout the whole keyrange, not in a sampled form mind you but in a more accurate mathmatical form. i dont know how it sounds but like i said, i just like the idea.

kinda fucks with your head when you think about it.

prophet wrote:

well, creating sound from light isnt anything new at all actually. infact its quite old now. anyone who messes with graphics might know that light in computers is actually RGB (CYMK being darkness - print). From this light we are able to create images, ie. bmps. from bmps we are able to produce sound, ie. Fruityloops Beepmap. actually, Aphex twin made a sound on one of his albums using this kind of process. If you were to load the track up in say cool edit and view it in its spectral form you can actually see a face, which i think is his. but listening to it would just result in a weirdish kinda sound. which pretty much sums up his whole discography Rock on!

Yeah, I heard something mentioned about Mr. James using Metasynth or some such - wouldnt surprise me really cause Windowlicker sounds mental

The neural network synth idea sounds wild! Bet it would cost the earth tho.

I guess I was talking kinda about the 'big picture synthesizing' - if someone uses the chaos theory or quantum mechanics to to create sound - these processes(?) exist so they are evolution not revolution.

Sorry guys - I have a migrane for the last two days & I'm in work alone with my brain spinning. I think I'm having a panic attack !

Sorry - I'll shut up now - I wish I could go home.

Cheers,

whyterabbyt,

Creation of sound is part of physical science. You have to create sound by hitting, blowing, bowing, et. al., you have a sound source that resonanates a sound chamber in the acoustic world.

It's no different when you go to analog, you apply voltages to oscillators that have application of various subtractive and spectral processes which amplified becomes sound. The same with digital processes.

I completely agree regarding whether sound becomes something commercially useful but I doubt Varese or Cage would have cared.

I agree 'new' technology is old in terms of theory, but having the physical capability to do something and knowing it exists is almost like saying we can travel in ships that will get us between stars in a couple of years. According to theory it's possible but we are nowhere close to being able to do it. Similarly granular was discussed back in the 90's and probably before, I was at a symposium on 'future' music technologies when I heard about it.

But the question still is are these evolutionary or revolutionary processes in synthesis? My vote, evolutionary. Someone else might not care, great, that gives me a better chance at learning it first.

Prophet's discussion of neural networks is very similar to arguments I heard back in 1991 regarding implementation on granular synthesis. Light to sound although older than granular and available back in the early 90's is unique but as yet not beyond a very small range of people interested in experimentation. atmogen may change that, it's not too far off. And again, atmogen is evolutionary.

Creation of sound is part of physical science.

Ummm, no. Natural sounds are the result of the laws of physics, but not 'part' of it as such. Like salt isnt 'part' of chemical science and 'people' aren't part of biochemical science....

You have to create sound by hitting, blowing, bowing, et. al., you have a sound source that resonanates a sound chamber in the acoustic world.

For natural sounds, yes. But what you get is the result of radiation of physical interactions that happens to exist in the audio range of the spectrum.

It's no different when you go to analog, you apply voltages to oscillators that have application of various subtractive and spectral processes which amplified becomes sound. The same with digital processes.


I completely disagree. Its completely different.

The subcomponents used in an analogue synth may work as a result of the laws of physics,but doesnt mean that the audio generation process is any way related to the generation of acoustic sounds. Thats why they call it synthesis. Its very different, and off-handedly saying 'its no different' doesnt make your position very tenable.

Digital is different again. Just because some software synthesis models analogue or natural synthesis methods doesnt make it related. Just because there's a 'physics' basis for the operation of a transistor, that doesnt make digital synthesis 'physics-based' except in the very loosest sense of the phrase. You could do digital synthesis by hand on a piece of paper, if you were prepared to do the maths. Even less 'physics' involved.

I agree 'new' technology is old in terms of theory, but having the physical capability to do something and knowing it exists is almost like saying we can travel in ships that will get us between stars in a couple of years. According to theory it's possible but we are nowhere close to being able to do it.

Being able to do something, and having a reason to do that thing usefully are two different things. The latter is what I was talking about.

Similarly granular was discussed back in the 90's and probably before, I was at a symposium on 'future' music technologies when I heard about it.[/i]

Its way older than that.

But the question still is are these evolutionary or revolutionary processes in synthesis? My vote, evolutionary. Someone else might not care, great, that gives me a better chance at learning it first.

Evolving from what? Something had to come first. If the basis of what you're saying relies on your claim that analogue and digital synthesis are 'no different' from acoustic instruments, then how does a Moog 'evolve' from a piano?
Answer: it doesnt. Electronic subtractive synthesis was a revolution. Max Mathews' digital synthesis on a punched-tape IBM was a revolution. There have been a few major revolutions.

Prophet's discussion of neural networks is very similar to arguments I heard back in 1991 regarding implementation on granular synthesis.

And? Are you saying that because the arguments are the same, the techniques have to be related?

Light to sound although older than granular and available back in the early 90's is unique

If its unique, its a revolution, which you say doesnt happen.

but as yet not beyond a very small range of people interested in experimentation. atmogen may change that, it's not too far off. And again, atmogen is evolutionary.

Lets get this clear. Data conversion or analysis isnt synthesis. Atmogen is an additive synthesiser. Where the control data comes from is irrelevent to that; it remains an additive synthesiser whether it uses images, ADSR envelopes, or values derived from stocks and shares values.

Any ROMpler or synth that uses samples is evolutionary, even those “new” granular synths.

I consider FM as evolutionary since some of the principles of FM and AM really evolved from subtractive analog synths that used oscillators as modulators.

Additive synthesis evolved from organs.

Even subtractive synths were proceeded by the acoustical manipulation of natural instruments as people learned how to change the sound by filtering high frequencies by various amounts.

I would say that some synths are higher on the evolutionary scale than others, but I have a hard time thinking of anything really revolutionary.

Robert

This thread seems to be more about 'innovation' than 'revolution'. To me, a revolutionary VSTi would be one that made people stop using their old ones - not necessarily one that is completely novel in it's design.

I'm personally about to start a 6 month research project about information theory-based synthesis, involving a new concept I like to call 'auditory entropy'. We'll see if it leads to something interesting!

I tend to think of it as starting out with box with strings strung across that is plucked or hit with sticks. That leads to something like a harpsichord, piano, guitar, etc. Alot of the musical instruments were pretty 'general' for a long time. The church organ became grander and more mechanical - and finally the saxophone in the 19th Century (?) was this 'missing link' to the brass section. Through the 20th Century, the Martinot, the Sackbutt, RCA, Therimin were to me 'revolutionary' in that they not only emmulated say a clarinet, but could go beyond a clarinets limitations in modulations, pitch, etc. RCA revolutionised music composition with it's punch card approch to not only programming but performing music. Therimin totally remapped a players action. And of course, good ol' Bob Moog 'revolutionised' the affordability of synthesizers to a curious moderate income person, and not the budgeted scientists at Princeton.
Chowning revolutionised the emmulation process with FM blocks in the early 70's. Oberheim (I think it was?) with polyphony, and then EMU finally with the Emulator in the 80's. Would all are musical plug ins and $50 Casio's have happened if Ondes had said 'You know, I'd rather make a church organ the size of the vatican!', or Moog had said 'Naw, I'd rather market the MAXI Moog for $150,000.00 and make some real coin!', or the FCC banned Emulators for 7 years due to their copyright infringement potential (like they wound up doing with DAT - which revolutionised DCC, MD, MP3, CD-R, CD-RW, DVD-RW, etc.) And finally, if Synth Edit and soft synths were still in C language for us all? I certainly would not join the 'evolution'.

But if I can grab a gun that's cheap and loaded, I WILL JOIN THE REVOLUTION! Viva Steinberg! Long Live VST! Hardware, we don't need no stinking hardware! That is 'old leader' thinking.

That Hartman neural thingie, isn't that just analyzing the sample with FFT and re-synthesize it with additive synthesis? And perhaps sampling the remainder that the additive synthesis misses.

And the "neural" is just marketing hype/slightly different analyzing than done before?

I think so and if that is the case it has of course been done before.
So no revolution there i am afraid.

Of course it is done better/cheaper than before.

I believe one of the Kawai additives could do this with a external program.
And the Synclaviers (and possibly fairlights as well) did this.

And i know for a fact that Oberheim was experimenting with this in the early 90:s but it was never realised.

jupiter8 wrote:

That Hartman neural thingie, isn't that just analyzing the sample with FFT and re-synthesize it with additive synthesis?

All I know is that Stephan Sprenger it the main man behind the algo, and that guy is a freakin' DSP genius. He is also somewhat of a neural networks guru, so it might actually be something revolutionary - or at least innovative.

Neuron is cool somehow.

BUT: I heard the demos from the website and I'm VERY underwhelmed. I mean it's fun that hardware freaks can do such stuff with such a nice looking box/controller like Neuron. These things (at least from the sonic experience) have been done with Reaktor and similar progs (Kyma, Max/Msp, etc) long before this really expensive synth came to this world.
The mp3 demos sound kind of cheap for such a price tag. I can do most things I heard there using Reaktor.
BTW, Reaktor is evo/revolutionary in quite some ways. Hehe...

this article kind of fits in with the topic

http://slashdot.org/article.pl?sid=03/02/28/1540204

WR, I note your responses and enjoy discussing them but I'm trying hard not to post verbosely (failing, but trying). I'd enjoy a conversation privately and perhaps you or I could distill the discussion here to save everyone who doesn't want long posts.

My reasons for posting were mostly to get an idea of where respondants stood on this issue. Language is a tough thing to tackle and in recent threads I've seen it several times create bad feelings or misunderstandings. Also, by using mixed metaphors and premises and being a native English speaking person, I have caused WR to disagree with my perceptions - this is my mistake. Maybe later it will be discussed.

What we have so far:
1. music is a form of science but what kind and what rules apply?
2. everything is evolutionary
3. some things are revolutionary
4. #2 and #3 vary between writers
5. there is a difference between applied science and theory
6. there is a difference between viability and need

Is discussion like this important? Compared to world peace - no. As for how we view what we use to express ourselves while using synthesis in music, yes. The front line of popularity in what is 'musical' is through people who understand and do. Rather than be passive it might be better to embrace and support new ideas and keeping an open mind.

Its 90% evolution and 10% revolution. If you look back over synthesis history there are three or four revoltions and everything inbetween those is evolutionary. The whole VA buisiness has been evolutionary. I would guese that the revolutions have been Analog syntheseis, FM synthesis, sampling, and Physical modeling although the latter doesnt seem to have made much inpact yet.

The thing i get sick of is the Marketing Bulshit, every software developer and his dog has to think up some jingo name for what they do in their synth. And the absolute truth of the matter is the Revolutionry ideas are often worthy of patents and evoltionry ones arnt. Or somthing like that.

You can Trade Mark your marketing hype jingo but a Patent has to have to have more than 50% meat in the Pie.

chris

Definitely feel free to email me privately x_bruce....