[music-dsp] DSP vs analog circuits Digitally switched analog circuitry

Michael Gogins gogins at pipeline.com
Wed Oct 24 23:35:13 EDT 2001


It is dubious that ALL acoustical or analog electronic instruments will
some day be modeleable by software. There are several reasons for this.
One is that we don't understand the acoustical or even the electronic
instruments well enough to model them, and we may never. But of course
you abolustely need a good mathematical model to produce a software
simulation. Another is that, even within the scope of what we do
understand, the complexity is indeed daunting - think of all the 3 x 88
strings of the piano interacting with the hammers, the soundboard, the
frame, and then with each other through those couplings, and we know
enough about the ear to know that yes indeed we do hear just about
everything that's happening there. Another is that physical effects and
electronic effects can involve chaos in the mathematical sense.
Sometimes chaotic systems can be "parametrized" so that you can
substitute one value of an initial state for another without a great
difference in behavior; in other cases sensitive dependence upon initial
conditions indeed affects the evolution of the system quite drastically.
In such cases, it might not even be possible to measure the values of
the controlling variables accurately enough to model the evolution of
the system, even though a human player in intimate feedback with the
instrument can perform quite well (bowing, embouchere).

I have no doubt that the models will get closer and closer to the
reality, and that it's worth doing what it takes to get there. I'm not
so sanguine that educated listeners can be consistently fooled - ever.

My personal experience is that current models of SOME instruments are
quite good, and fool educated listeners sometimes or for short periods
(usually simple wind or brass instruments, or mallets). Guitars,
harpsichords, they sound like the real thing but don't really fool one.
As for pianos and strings, forget it, no way. Not even close.

As far as I'm concerned, this is all barking up the wrong tree. What
sounds can one make with software that can't be made with acoustical or
electronic instruments? What kinds of control can you get with software
that you can't get with a skilled player or a sequencer?

-----Original Message-----
From: owner-music-dsp at shoko.calarts.edu
[mailto:owner-music-dsp at shoko.calarts.edu] On Behalf Of Christopher
Weare
Sent: Wednesday, October 24, 2001 6:11 PM
To: music-dsp at shoko.calarts.edu
Subject: RE: [music-dsp] DSP vs analog circuits Digitally switched
analog circuitry


Actually, our understanding of physics is sufficient in most cases to
completely model any analog system mathematically.  The problem comes in
generating accurate numbers to represent the state of a system at any
given time.  Given unlimited precision of measurement then yes, you can
specify completely the state of a system at any given time.
Unfortunately, unlimited precision is rather impractical.

And here in lies the problem.  For many "simple" analog systems it is
just too damn complicated in this day and age to accurately simulate the
system in the digital realm.  The Guitar-Amp-Air-Player system being one
that springs to mind.  We can do a decent job these days.  I love my
vg-88 for instance, but it aint the real thang.  We still have a ways to
go.  I am confident that given enough processing power we will be able
to simulate any analog system to sufficient accuracy that no one will be
able to tell the difference in a black box comparison.

That day is yet a ways off.
-chris


-----Original Message-----
From: Garth Brantley [mailto:garthhudson at yahoo.com] 
Sent: Wednesday, October 24, 2001 2:36 PM
To: music-dsp at shoko.calarts.edu
Subject: Re: [music-dsp] DSP vs analog circuits Digitally switched
analog circuitry

   Agreed that there are many advantages to digital. 
IE saving your setup easily, transporting your
algorithm accross the internet, less noise, etc.
    That said, with dsp you can only ever implement a mathematical
model.  Analog circuity goes beyond any mathematical model that tries to
describe it.  No mathematical model can ever completely sum up the real
world.  Godel has proven this already.
    From where I sit, the thing that seems to drive
people to make & listen to music is its ability to
transcend the world of conscious ideas.  IE to
transcend mathematical models in to the world of the unconcious.
    The idea of the realtime feedback loop seems
central to music.  The act of musical performance can
be described as a feedback loop between your muscles &
your brain.  In fact, this feedback seems to operate
at a speed much greater than the resolution of
conciousness.  The process musical training can be
seen as training your brain to get very tiny amounts
of conscious control over that feedback loop.
    To me the above goes a long way toward explaining
the importance of the vacuum tube and the transistor
in music making in the 20th century.  Anyway, that is
just my 2 cents.

 -Garth Brantley

--- Sampo Syreeni <decoy at iki.fi> wrote:
> On Wed, 24 Oct 2001, Garth Brantley wrote:
> 
> >I agree wholeheartedly that we need to have some
> more digital/analog
> >hybrid stuff.
> 
> I'm not saying that analog doesn't have its place in
> the signal processing
> world, but I do find most propositions for
> digital/analog hybrids a bit
> misguided.
> 
> If one can demonstrate that a given sound cannot be efficiently 
> produced by digital methods alone, I'm all for the
> incorporation of analog
> processing. However, most people who advocate analog
> processing seem to
> think that there is something very special about it
> when compared to
> digital. This is not so. Most things achievable in
> analog circuitry can be
> efficiently implemented digitally. In fact,
> *everything* one can formulate
> mathematically, one can also transform into the
> numeric domain, albeit
> sometimes with considerable computational
> complexity.
> 
> As the digital domain is much more configurable,
> suffers less from thermal
> noise, and processing power keeps growing steadily,
> I would *not*
> recommend analog implementation for anything but the
> most twisted "analog
> sounds".
> 
> >I can see an analog audio processing board you
> could place in a PCI
> >slot. It would have opamps, simple components, some
> transistors (if you
> >use an external box, hell through some 12ax7 tubes
> in there).
> 
> If a tube sound is what you're after, why not model
> your l2ax7 in
> software?
> 
> >All the analog components could connected together
> to form different
> >circuits with software.
> 
> The trouble is, what do you do when you want to
> share what you've
> achieved? The crux of the matter is that most of the
> analog functionality
> can be implemented digitally while the converse does
> not hold. When you
> consider that most people have a PC nowadays, there
> is a big difference
> between relying on a proprietary analog board "for
> that special sound" and
> constructing your algorithm in a form intrinsically transportable over

> the Internet.
> 
> Sampo Syreeni, aka decoy - mailto:decoy at iki.fi, tel:+358-50-5756111
> student/math+cs/helsinki university,
> http://www.iki.fi/~decoy/front
> openpgp: 050985C2/025E D175 ABE5 027C 9494 EEB0 E090
> 8BA9 0509 85C2
> 
> 
> dupswapdrop -- the music-dsp mailing list and
> website: subscription info,
> FAQ, source code archive, list archive, book
> reviews, dsp links
> http://shoko.calarts.edu/musicdsp/
> 


__________________________________________________
Do You Yahoo!?
Make a great connection at Yahoo! Personals. http://personals.yahoo.com

dupswapdrop -- the music-dsp mailing list and website: subscription
info, FAQ, source code archive, list archive, book reviews, dsp links
http://shoko.calarts.edu/musicdsp/


dupswapdrop -- the music-dsp mailing list and website: subscription
info, FAQ, source code archive, list archive, book reviews, dsp links
http://shoko.calarts.edu/musicdsp/


dupswapdrop -- the music-dsp mailing list and website: subscription info,
FAQ, source code archive, list archive, book reviews, dsp links
http://shoko.calarts.edu/musicdsp/




More information about the music-dsp mailing list