[sdiy] Re: FS-2

[Magnus Danielson]

From: "jhaible" <jhaible at debitel.net>
Subject: FS-2 (was: Re: [sdiy] analogue DX-7-style Phase Modulation using quadraturesignals ?)
Date: Sun, 26 Oct 2003 19:52:48 +0100

> Hi Magnus,

Hi Jürgen,

> > This is where I chime in and start to question your statement. In essense I
> > think your saying "it's not reasnoble" here, and that is what I think is
> > wrong,
> 
> You're probably right. It wasn't my intention to disqualify any method
> of making analogue PM - I just remembered a recent discussion on synthdiy
> where my concluding impression was than no one had a really good solution.

Ah! Right. I wanted to make a fairly quickly "brain dump" to give an overview
of the many possible solutions that I have come up with. I'm sure there are
many more. I brougth them up to make an few counter-examples.

> Now I know Don, and you too, are working on it - so much the better!
> I'm looking forward to see what comes out of it.

Well, I think we have two different topics here:

1) Generic phase modulation (i.e. PM on normal/generic waveforms)
2) Sine/Cosine phase modulation

The first one is what I have been concentrating on, while the later is what you
and Don seems to have been working on.

> > So, when will we have an analog DX-7 clone? ;O)
> 
> This, and the entire PM matter is not in my own focus right now.
> The whole idea came from my work on a new Frequency Shifter, and
> unlike my fist one, where all is hard-wired, I want to make an open system
> where you can patch everything to everything, and I want to find out
> if that PM application would be worth to implement in form of a sin / cos
> network for arbitrary CVs rather than quadrature VCO only.

Hmm... just WHY I am thinking about the Serge FS? I used the oscillator from
that one to do the modulation on the phaser module (I needed a ssssslllloooowww
wave).

> BTW, I've finally found a reason to really have 2 Dome filters, with the
> possibility
> to have an external (non-quadrature) VCO replacing the internal QVCO!

Right.

> The downside, which I often said, is that the second dome filter does not
> allow near-zero shifts. BUT an external (or internal, but non-quadrature),
> well tracking VCO + 2nd dome filter allows precise V/Oct control of the shift
> frequency, und thus perfectly harmonic playing with altered spectra!

Interesting. Yes, I never rulled out this alternative myself.

Oh, you've just reminded me that I don't have a frequency-shifter!!! Yeat I
seems to spend so much time on learning about them. HMMM!!!!! Is this "know
your enemy"?

> So instead of building a "super thru-zero-quadrature" VCO that also tracks
> precisely, I rather use *two* VCOs, one for thru-zero and ultra-slow
> modulation, and another one, with precise tracking, and a tracking VCF
> for sine purity, for the harmonic stuff.

Maybe keep them locked in the intermediary region and let the user get the
best performance depending on where he/she put is working for the moment?
That is, naturally, if the control interfaces of the VCOs is compatible
enought, else it breaks up. Just an idea...

> I might just use a MOTM VCO + 440VCF, the two dome filters and two
> multipliers to get precise, "harmonic" frequency shifting - and use a simple
> QVCO for the slow phaser-like stuff.

Right... but phase modulation behaves differently... so it might still be of
interest at higher frequencies.

> Sorry that the thread is going into a different direction for me - but this
> is the background why I was asking - to find another application of the thing
> I'm just building, and keeping it "open" for such applications.

Actually, I think that this thinking is a good one and as I stated above, it is
really one of two different threads. I intend to follow both of them with
interst I migth add.

> The other, actual PM thread, is very interesting as well. Yes, Sines run thru
> Wavefolders certainly have that "Bessel spectrum" sound of FM synthesizers.
> I noticed this before, but I never thought that it could be actual PM that's
> going on there. So that postprocessing - PM stuff that you, and Roman,
> mentioned is very interesting as well!

Indeed! You see, I've long known that PM is quite interesting and if you think
of it, you know that I spent some time to explain other properties in form of
phase modulation. This both in analysis and modulation. Figuring out how to do
various forms of phase modulations have those come as a natural thing.

For instance, PWM introduces PM in addition to the pure PWM stuff. We know how
to do PWM with (sawtooth based) and without (triangle based) but what if we
want the amount tweakable or possibly even independent of the actual PWM
waveform? What if we want it without the PWM to start with?

To some degree I think that is ortogonal with more elaborate variants of
frequency shiftings. I think it is natural that you dig where you stand, which
in this case is in the frequency shifter end of things. I have yeat to spend
any time over those specific problems.

Cheers,
Magnus