[sdiy] Re: Taking pre-orders for JH. 36dB/oct LPF now

Andre Majorel aym-htnys at teaser.fr
Sun May 31 20:13:20 CEST 2009


On 2009-05-28 08:45 +0200, JH. wrote:

> I've ordered the boards from the factory already, but I always
> get more than I've ordered (and have to pay them: small print of
> contract that I have to take over-production),

Interesting commercial practice.

> >Personally, I'm not sure about this modulation source. I suspect
> >that, on my unit, the waveform pot would be permanently stuck in
> >pure sine position, making the scanner redundant.
> 
> I am *very* sure about this modulation source, because I came to
> love it in my JH-4 synthesizer module over many years.
> Don't under-estimate the Sample&Hold modulation on a phaser with
> high resonance! And this is a very special S&H, with a kind of
> overshot or spike (for lack of better description) at the
> beginning of each new random sample, which makes the S&H more
> "agressive" than a "perfect" S&H would be.
> But there is more:
> The waveform crossfading ("scanner") is arranged ina way that
> S&H is adjoining "square" on the left side, and
> "self-modulation" on the right side. That means, in S&H position
> you have true random staircases. Turn the waveform pot slightly
> to the left, and you get a tad of regular rectangle mixed in:
> the steps become less random in a way that large steps are more
> likely. Turn the waveform pot slightly to the right, and the
> random S&H gets mixed with the dirt of an audio-rate
> self-modulated filter. That's especially impressive at high
> resonance settings. I got the idea for this from the Prophet 5's
> Poly-Modulation and the CS-50's and Minimoog's external
> modulation input, and - for a phaser application - from the EMS
> Synthi HiFli.

Thanks for the explanations. Where does the self modulation comes
from ? lfo_mod seems to be in the wrong direction. Perhaps AUX
MOD IN should be normalled to something else than ground ?

> >The other sources I see myself using (velocity, EG and sawtooth
> >LFO) all have to come from other modules anyway.
> 
> Yes, it makes a lot of sense to use external modualtion sources.
> My on-board modulation source is kind of special (I hope!), but
> it doesn't have saw waveforms, for instance.

Seen from here, it's plenty special. :-)

> >In theory, each pole could individually be switched between
> >all-pass (phaser) and low-pass (filter) under voltage control. If
> >all poles are sonically interchangeable, a single CV could be
> >used with the number of LP poles being 6/5 of the CV. No idea how
> >to do that, nor how useful it would be.
> 
> I'm not entirely sure if I understand what you mean.

This is what I had in mind :

CV		Pole 1	Pole 2	Pole 3	Pole 4	Pole 5	Pole 6
       ~ 0.42 V	LP	LP	LP	LP	LP	LP
0.42 V ~ 1.25 V	LP	LP	LP	LP	LP	AP
1.25 V ~ 2.08 V	LP	LP	LP	LP	AP	AP
2.08 V ~ 2.92 V	LP	LP	LP	AP	AP	AP
2.92 V ~ 3.75 V	LP	LP	AP	AP	AP	AP
3.75 V ~ 4.58 V	LP	AP	AP	AP	AP	AP
4.58 V ~       	AP	AP	AP	AP	AP	AP

The idea is that the higher the CV, the more high frequency
content there is.

Looks like it could be done with an LM3914-type circuit.

> But the 6 stages are interchangeable only with the "Slope" (or
> pole spread) potentiometer at ccw position. When the poles are
> spread apart, it makes a difference which one gets the
> complementary zero (= switchet to all pass), and which one
> doesn't.

Okay, so there are not 7 but 2**6 = 64 combinations. I can't think
of a way to map this six-dimensional space to a scalar. Not one
that would be acoustically intuitive, anyway.

Ordering the combinations by amount of HF energy is not
straightforward because a combination that has more HF than
another when the first pole is at 100 Hz might have less when it's
at 5 kHz or vice-versa.

Ditto for ordering by response at 20 kHz (or any other fixed
frequency).

-- 
André Majorel <URL:http://www.teaser.fr/~amajorel/>



More information about the Synth-diy mailing list