[sdiy] Taking a Step towards - - --((FUTURE-PREDICTIONS))-- - -

[Magnus Danielson]

From: "Paul Maddox" <P.Maddox at signal.qinetiq.com>
Subject: Re: [sdiy] Taking a Step towards - - --((FUTURE-PREDICTIONS))-- - -
Date: Wed, 14 Jan 2004 10:07:18 -0000
Message-ID: <01a701c3da86$327fb950$a45e23c0 at dra.hmg.gb>

> Magnus,

Paul,

> >I am not going to debate weither analog or digital is THE best. IMHO they
> have
> >different strengths and weaknesses and often it also boils down to the
> >engineering details of either world, many times it is even "best current
> >practice" which doesn't really is the best, just the "current practice" and
> >nothing else.
> 
> agreed, but a lot of people seem stuck on 'analogue is best and everything
> else is just rubbish'

Yup! Its a simple way of looking at things, but I don't think about something
just because it is simple, sometimes I do it because it is hard. Sometimes I
even find that it isn't hard, it is just that it is unusual (I love doing the
unusual).

> > Hopefully we can help improve the state by sharing ideas
> 
> indeed, me too, The MonowaveII Code goes up on my website tonight for those
> into DSP programming.

Ah!

> >and Ithink I've seen many very interesting advances on the analog side from
> (other)
> >members of Synth-DIY (Jürgen, Ian, René and all those which my poor mind
> just
> >forgot the names - please forgive me, but you've all contributed one way or
> >another).
> 
> I'm not disputing this, but 99% of this list geared towards analogue and
> seems to have a passionate dislike to digital, without looking at the whole
> picture..

Just because something is not being discussed alot here doesn't in any way
exclude the possibility that it could be. We have had very interesting
discussions on stuff which is really into the digital domain and I can't
recall that it was being taken badly.

> > Wavetables is to some degree the poor-mans sampler,
> 
> I would disagree.

I say that since you CAN use it to emulate sample-players. Sample a waveform
and transform it into wavetable sweeps. It is one way to approach the
wavetable-stuff, but not necessarilly giving the truth or a good idea of what
it is about, but you can view it that way... too.

> >but then again is a wavetable-sweep a quite a different approach to that of
> >sampling, sample-stretching etc.
> 
> yes, VERY different.

Well... yes and no. Look, now you are the one trying to draw strict sharp lines
between two things... ;O)

My point by saying stuff like that is to bend the subject all over and bring up
a spectrum of different concepts and show how they relate with various
intermediary steps.

> >It is certainly subjective and people also listens to different
> characteristics
> >and have different preferences. Some like a big fat Moog-ish sound at all
> times
> >and others want thin sounds, cracking sounds or whatever and your
> preference
> >will certainly become a big judgement on what is "right" and what is not.
> 
> bingo, like I said, everything doesn't HAVE to be analogue, its a matter of
> choice..

Yeap, and I also try to separate the technology, what it acheives and what we
judge it by.

> >IMHO is the whole digital vs. analog discussion old and booring.
> 
> I agree, and I tried hard to avoid it.

Yeap, and here you sit in the shit again... ;O)

> >If somebodycomes and say "I really like this feature of this system" then
> that's more
> >worthfull.
> 
> for me this is the Chameleon, its standalone (no vst plugin), its powerfull
> and flexible, its reprogrammable and whats more I can write stuff for it, no
> soldering, no wasted PCBs, no mod wires, no burnt fingers!

WHAT? Are you going to robb us from all the fun, necessities and hurdles of the
perfect sound???? GHAAAWWWW! Return to thy dungeon thou evil dragon!

> >What happends when you combine the analog and digital world in interesting
> >ways?
> 
> I dunno, closest I've seen is things like the PPG Wave, Monowave, Prophet
> VS, evolver and so on..

Those are still conservative designs, except for the evolver which actually was
one of the first things which gave me a thrill when reading about. The evolver
has a much less traditional view on stuff and tries to do stuff in digital
which is good in digital and vice versa. But, can you take it further?

> >For instance, some time back I proposed a new way of building a analog-
> >controlled (CV!) digital delay with analog input and output (could just as
> >well been digital input and output). In there I used a digital storing
> through
> >DRAM or whatever but under analog control. I've never seen anything like it
> >before (which doesn't mean it hasn't been done). Really just an
> >BBD-replacement (which was the point anyway).
> 
> its do-able, but again, most modular people seem to balk at the idea of
> anything digital in their system.

Yeah, yeah, yeah... I DONT CARE!

> >So, my point is... try to solve your problems in the best method possible
> and
> >learn what issues is important and which isn't. There is too much tradition
> >and folk-lore going around here and some of it doesn't make sense. Yes,
> there
> >is an effect which we can attribute "warmth" - I can hear it - but I still
> >think there is a little too much mumbo-jumbo going on around it which is
> part
> >of folk-lore more than a scientific way of looking at it. The real quest
> there
> >is to know it well enought so we can put a knob on an oscillator (analog or
> >digital) and control it from "cold" to "warm" as we like it.
> 
> agreed, this would be storming, I've seem attempts using noise and so on,
> but there's something other than just pitch/amplitude/dcoffset changes going
> on.

Characterisation, we need characterisation!

> >If you want imperfections - just look at these bloody speakers we keep
> using!
> 
> <ROFLMAO>

Yeap! They are worth laughing at.

> >Why not?
> >(Said in a VERY retoric fashion and also honestly asking at the same time.)
> 
> part of the quoted reason for 'analogue' warmth, at least the ones I get, is
> that it isn't 8bit, 16bit or 24 bit, its infinate resoloution.

I haven't heard that one. That the "tagginess" of quantization doesn't go down
well with people is certainly not news.

What about cold analogue gear?

> >Have you tried? All you have to do is get the sawtooth core up to spec.
> >I clocked mine up to 160 kHz. You can do more or less ugly frequency
> >multiplication tricks as well (see my webpage).
> 
> I've tried..
> 160Khz just inst high enough... that would only give you max pitch of
> 1.25Khz using PPG wavetables..
> if you want the full 8khz, 128*8000 = 1.024Ghz

Um... you mean 1,024 MHz or PPG where doing serious magic back then... (knows
better).

Anyway, yes... 1 MHz seems high. There is a neat app in the CA 3080 datasheet
showing how you can make a continous sweep from 1 Hz to 1 MHz using a pair of
CA 3080 and a CA3160. The frequency control is at one of the Iabc inputs so a
standard expo-curcuit using PNP transistors (see ASM-1 VCF for instance) should
allow for a fairly straight-forward way of acheiving a full-range analog
oscillator to control a PPG-style wavetable. So, what's the problem? Known
components and not too uncommon techniques. Well, high-freq tracking is just
not being explored.

Anyway, there is a path there. It hasn't been walked for quite some time, but
it's there. I'm doing this partly to annoy you, but partly to show my point
that we keep thinking in a little too much the same fashion all the time.

> > Actually, as I recall it, this is really from the point of view that the
> PPG
> > concept originated from. The PPG Wave 2.3 is just a continuation off a
> line of
> > developments from the traditional oscillator. The waveshaper went through
> > various transformations before the oscillator core was changed into its
> digital
> > equalent with all the obvious improvements that had. Then the waveshape
> > improvements continued.
> 
> yep, the 360, for example, has no filter, it sounds great, but musicians
> wanted a filter, so palm made the wave 2.0, with a filter. People like what
> they are familiar with.

Indeed. And then there where PPG gear before the 360.

> >Actually, I think this argument starts of in the wrong end from both sides.
> >I think there is a number of usefull techniques around. Some is dominant in
> the
> >analog world and some is dominant in the digital world. They acheive
> different
> >things amd have different pros and cons. It may initially look like they
> are
> >compound properties of one or the other technology, but I beg to differ. It
> is
> >not at all as clear-cut as people make it, and you can learn alot by trying
> to
> >force a technical solution over border. Hell, people are so locked up in
> their
> >way of thinking that it is really annoying!!!
> 
> agreed.. and very well phrased.

Thanks!

> > Let's face it, more things is happening in the digital domain right now
> because
> > that's where the market is and prices of hardware is dropping (for a
> certain
> > functionality) due to the development in semiconductors and digital stuff
> is on
> > the forefront of that while much analog stuff is still about where it was
> in
> > the late 70thies except for a few improved components (like op-amps).
> 
> its also cheaper to develop, see above comments about PCBs.

It's dirt-cheap today. We can spend much more resources in processing on a
single function than in the early days. Today it is really just the imagination
which puts the limit except for a few SERIOUSLY powerhungry things.

IMHO one error is that a too low sampling frequency is being used.

> >Actually, they are not only grey, they are different colours of grey!
> 
> :-P
> Trust you!

You better, you better... ;O)

Cheers,
Magnus - who actually had a good day at work - thinking differently ;O)