[sdiy] ... Simulating a Moog

[Richard Wentk]

At 16:17 03/05/2004 -0700, Metrophage wrote:
>--- Richard Wentk <richard at skydancer.com> wrote:
> > But softsynths win easily on reliability, polyphony, versatility,
> > cost, size, and ease of transportation.
>
>I'd concede on most of these points, but certainly not that of
>reliability

I suspect homebrew hardware gives the illusion of being more reliable 
because any repairs appear to be folded into the construction phase. :-)

But things like big Moogs and Prophets can go wrong a lot. And getting 
spares can be hell...

>, and versatility is relative to how readily ones
>applications can be translated to another medium.

Or more probably how easy it is to realise your ideas. Analogue is really 
quite limited sonically. That's not a bad thing, it's just that it only 
does one thing really well - sounding like itself.

>Special purpose hardware is always, i think, more reliable than general
>purpose hardware. The size and especially cost is typically the issue
>which prompts adapting to general-purpose hardware. Then there is the
>issue of how long one needs to wait for the 'faithful replica'... When
>is the convincing VST plugin of the Buchla 200 due? How about Eric
>Barbour's 'Phattytron'? I could expire before those are ever modeled.

Most people are going to expire before they get their hands on that kind of 
hardware too. :-)

>I have been playing with Max for years also. Quite handy for making
>sequencers on my mac. But for many applications i need to make my own
>externals or plugins. So what then is the difference between taking up
>programming to treat the computer as the mathematical, logical
>processor that it is? As i begin to do so, i don't have time to waste
>duplicating material instruments.

But that's the whole point of something like Open Source. Projects like the 
ASM1 are sort of (very kind of) the hardware equivalent, but building and 
buying hardware is sloooooow.

Is it really such a stretch to see a group like this collaborating on an 
Open Synth collaborative DSP project ten years from now?

>I think emulation is always a waste of time and energy. Using digital
>to re-animate analog circuitry is as foolish as using synths to
>'emulate' acoustic instruments such as oboe, tuba, violin, etc. The
>only reasons to try this are ;familiarity, and the fact that most
>people don't understand or like the truly digital functions and
>esthetics.

That's true up to a point. But that's a function of the versatility of 
digital. It's creatively more challenging to deal with an open system like 
Reaktor, and more so still with Max/MSP, than with the relatively simple 
functional units and traditional patchings possible with a modular.

I don't think modulars *have* to be simple, but there seems to less 
interest in creating new modules and functions. E.g. instead of a 
sequencer, how about a parallel step memory that remembers dynamic inputs 
instead of just reproducing fixed knob settingS? And keeps them in memory 
when powered down? There's a ton of possible applications like that, but on 
the whole analogue design seems to be a somewhat retrospective art.

(I'll admit I got excited when I saw someone asking for a Spectrum Memory 
because I thought that sounded like a fun module. ;-) )

>It's just easier to sell stuff that is familiar, rather than explain a
>new kind of music. I have been playing with VST plugins for six years,
>and people just barely enable real digital synthesis, like a Kontakt
>pitch stretch, or reducing bitrate. As for modular digital
>environments... my computer is to slow to use them for crunchy DSP.
>Still cheeper to make a standalone DSP board and program it than to buy
>a new (general purpose) computer system.

I rarely run out of steam with Reaktor, and finding out just how far you 
can go without compromising sound quality but still using maybe 10% of 
processor power is part of its fascination.

And it's not like you have to do the hard stuff with it. *Even if you only 
think in analogue terms* it can still do a lot. E.g. combine a step memory 
with a vocoder and you have a very interesting module.

>As for control, unless you know exactly which knobs and switches you'll
>need when- it still doesn't hurt to have a room full of specific
>controls.

It does if you have to move it. :-)

Richard