A CHORUS is essentially the same as a FLANGER. The main difference is
the feedback (also known as "Regeneration" or "regen" on some flanger
boxes). So if you add a feedback path to a chorus, you get a flanger.
Usually the "sweep" range (i.e. amount of BBD clock modulation via LFO)
is bigger on a flanger than on a chorus; in the range of 20 ~ 50 msec
compared to 5 ~ 8 (?) msec.
Additionally the "delay" setting (fixed delay amount of the delay line)
is longer on the flanger, but the delay setting is not critical to the
sound.
The resulting block diagram for a chorus/flanger unit is like this:
"speed" "delay"
-\/\/\-> -\/\/\->
| ^ |
--------- \ ---------
| LFO |---/---| clock |---
--------- \ --------- |
mod. intensity | ^ feedback
("sweep") | / ("regen")
-----------------<---|----\-----
| | / |
| |\ V | ^ effect
|---| \ -------------- | / depth |\
| >-------| delay line |--+-\-+---------| \
in o------+---| / -------------- / | | >--o out
| |/ | ---| /
| | | |/
| | |
| | |\ | |\
| |-| o-----| \ inv
| |/ | | >--o out
|---------------->------------------------+--| /
|/
The NE570 thing is a different story. They build with two NE570 (or
similar) a companding system. The delay line usually is (in equipment
from this era) an analog delay, so called "BBD" (Bridged Bucked Delay).
This can be imagined as a series of capacitors which hand over a
certain amount of electrical charge (refelcts input voltage) with
every clock cycle. The (analogue) input signal stays analog in terms
of level, but becomes quantized (=discrete steps) in terms of time
progression. So there is quite an amount of noise generated in such a
BBD line. (Additionally there is of course the "sample" frequency
present in the output signal, which must be filtered. This is exactly
like in digital systems. That's why you get aliasing effects when
choosing the clock frequency too low). To improve the overall S/N
ratio they make this compander stuff. NE570 is (or was then) one of
the cheapest and easiest solutions for this.
The "effect depth" control is not very useful, as the chorus or flanger
depth is controlled by the "modulation intensity" control.
So the most important controls are "(modulation) speed" and
"(modulation) intensity" for a chorus device and additionally
"feedback" or "regen" for a flanger. If you look at guitar effect
pedals, this is what you get there.
In case of a programmable synthesizer these controls should be
preferably memorized with each path. I dont know about the Poly800
but I can tell from other synths that this will most probably need
a firmware change...
A PHASER is again very similar. The main difference to the
chorus/flanger is that there is a (constant) phase shifter instead of
the (constant) time delay line. This phase shifter is constructed out of
an allpass filter, of which the corner frequncy is varied by the LFO.
You get usable phase shifts with a 6 ~ 8 stage allpass filter. The
phaser may or may not have a feedback path. (I am not sure now. Probably
a phaser with feedback would sound similar to a flanger?).
Anyway, in both variations (delay/phase shift) a comb filter like
frequency response is achieved by summing the delayed or shifted signal
to the dry signal, which is even mor enhanced by the feedback. The
(constant) delay gives zeroes in the frequency response in equal deltas,
e.g. at 1kHz, 2kHz, 3kHz,... whereas the (constant) phase shift results
in zeroes which are in equal intervals (1kHz, 2kHz, 4kHz, ...).
The (slightly) varied delay of a chorus actually generates a vibrato
(modulation of pitch) in the effect-only signal. When this is added to
the dry signal this simulates an (strings) orchestra or choir where each
single instrument is not in technically perfect (i.e. locked) tune with
each other and so adds some movement to played notes.
I hope, I am mainly correct with this. And I apologize for the long
mail!
Johannes
jo.synth@...
korgpolyex@yahoogroups.com wrote:
> I tried implementing a delay and depth control, with mixed results. After
> the sine generator there is an opamp buffer cicuit. I replaced one resistor
> in the feedback loop to change the depth - works fine, but doesn't quite
> get the range nessessary. I'll have to play with it a bit more. As for the
> delay, there are two resistors forming a divider connected to the + input
> of the opamp. I connected another 100k knob in here.
>
> The delay control proves useless. I could probly do it better in a different
> way, but the sound of the delay line being underclocked is just BAD. No
> distortion, no interestingly brutal signal degredation - just a high pitched
> buzzing sound. I've only got room for three knobs with the way I'm laying them
> out, so this one will probly be replaced, and maybe the depth as well should
> it prove less interesting then wet/dry mix and feedback.
>
> As for wet/dry and feedback, how would you go about implementing them? There
> appears to be some form of raw mixture between the original signal and the
> delayed one near the end, I suppose I could simply add another knob in here..
> and then for the feedback I'm clueless. The basic signal path of the chorus
> is as follows: Ne570 stage one, buffer, delay IC, buffer, 570 stage two. Where
> would you mix the signal back in? I'd assume the signal would be taken off the
> second stage of the Ne570, right before it's mixed with the original, or just
> after the second buffer. And ideas?
>
>
