MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

Oscillators, LFOs and EGs with bounds modulation
MUFF WIGGLER Forum Index -> Eurorack Modules Goto page 1, 2  Next [all]
Author Oscillators, LFOs and EGs with bounds modulation
pugix
Have you heard of bounds modulation? Bounds are the lower and upper limits of voltage excursions. A typical VCO has fixed bounds, often +/- 5 volts (10 volts peak-to-peak). The peak is a boundary. Similar for envelope generators, which typically have unipolar output from zero to some (usually fixed) upper bound.

Peter Blasser designed 'bounds modulation' into his Eurorack VCOs, Denum and Fourses. These triangle oscillators let you modulate the bounds. With Denum you can only modulate both bounds at once, but with Fourses, each oscillator has CV inputs for both upper and lower bounds. At audio rate bounds modulation, the Denum changes both amplitude and frequency at the same time.

I've discovered a different type of bounds modulation on the Starling Via META module. The META has Oscillator, EG, and LFO modes (among others). What makes META unique to me is that the upper and lower bounds have CV input, as well as knobs for manual setting, attenuation, etc.. But, contrary to the Blasser designs, modulating the bounds doesn't change the frequency, but only the amplitude. AM over your oscillator is like having a VCA or ring modulator patched after it. With the META, the approach is reversed. The bounds of the output itself can be modulated independently.

I'm wondering if there are other (mainly Eurorack) modules that have independent CV over upper and lower bounds. Just including a VCA on the output doesn't do it, because both bounds are equally affected by a VCA.

Also, I'm wondering if bounds modulation in general offers any really new spectral dynamic possibilities, or whether it's just a different approach to what can be done using standard modular patches.

Thanks,
authorless
I don't have either module so I don't know for sure. It sounds like either attenuation and offset or min/max.
Navs
^ that was my thought too: using a min/max as a voltage limiter.

https://navsmodularlab.blogspot.com/2011/08/patch-tips16-analogue-logi c.html

The effect sounds like clipping; maybe that's another method you could use.
mosorensen
The (quickly discontinued) Erica Synths Pico random module, called "Mask", also has this kind of bound adjustment. It is an unfamiliar concept, and I am unsure how to patch with it and how useful it is in practice.
peripatitis
a half wave rectifier and two vca's should do the work. no?
dooj88
peripatitis wrote:
a half wave rectifier and two vca's should do the work. no?


aka, blinds.
pieter
dooj88 wrote:
peripatitis wrote:
a half wave rectifier and two vca's should do the work. no?


aka, blinds.


How does Blinds work as a halfwave rectifier?
pugix
Min/max doesn't change only bounds; it slices off the tops and bottoms of waveforms, changing their shape as well as the bounds.

Neither does rectification, which only removes negative going voltage, setting the lower bound to zero.

A simple VCA can reduce both bounds of any signal.

But what META can do is to take a signal going +/-5V and lower just the upper bound, say, so the signal becomes +4/-5V, or both bounds, so it can swing between -3V and +2V. All the while, if the signal is an audio oscillation, the pitch doesn't change. And the modulations can be other audio frequency waveforms. So here is a type of AM, it seems, that is like an asymmetrical VCA or ring modulator.

I'm excited to explore the sounds that I can get from using this type of modulation on the META, for which there doesn't seem to be an equivalent in any other module.

Check out this description of the cross-fader in Via.

https://starling.space/via/platform-info#crossfader
electricfence
What about this: Take the output of a VCO and mult it, and then run each copy into it's own bipolar VCA, adjusting the amplitude as desired. Then run each of those signals into a rectifier to remove the negative portion of the signal, and then invert one signal, and finally sum the two signals together. In that case, wouldn't the VCAs give you control over each boundary separately?
pugix
electricfence wrote:
What about this: Take the output of a VCO and mult it, and then run each copy into it's own bipolar VCA, adjusting the amplitude as desired. Then run each of those signals into a rectifier to remove the negative portion of the signal, and then invert one signal, and finally sum the two signals together. In that case, wouldn't the VCAs give you control over each boundary separately?


Take a look again at the Starling crossfader block diagram.

https://starling.space/via/platform-info#crossfader

You can find any number of crossfader modules in Eurorack. As long as it is DC-coupled, why wouldn't an ordinary crossfader work? You would have to invert your thinking about it a little. If you crossfade two DC offsets at audio rate, the result would be a signal with the same frequency as the CV with offsets becoming the bounds. Instead of offsets, the inputs could be audio. It's probably not that unusual to crossfade two audio signals with another audio signal.

The META is basically a crossfader with a built-in contour generator that can also oscillate.
windspirit
Hey I posted something about this but apparently it didnt work.

As you mentioned min/ max/ rectifier will not work. However you should be able to patch this with a slew limiter and a window comparator. Either that or a slew and a pair of comparators and some logic/ flip flop/ clock divider sort of situation.
dooj88
pieter wrote:
dooj88 wrote:
peripatitis wrote:
a half wave rectifier and two vca's should do the work. no?


aka, blinds.


How does Blinds work as a halfwave rectifier?


perhaps i've misunderstood the request, but it sounds like if you were to mult the signal and add a positive offset to one input and a negative to the second channel using the attenuverter, modulating the 'IN' jacks would move the bounds of the attenuverted inputs on their respective channels. i suppose this only works if you have unipolar cv in the first place, however. this can be done with the other 2 channels of blinds.. attenuvert and then add an offset. not ideal, but it'll do it.

or maybe this is all jibberish... i've taken a several months off the ol bleebinbloops box, and i'm feeling a bit rusty getting back into it.
flashheart
OK, how about this...
A = signal VCA'd by Upper bound CV
B = signal VCA'd by Lower bound CV
Crossfade between A & B using input signal.
At +ve signal max you will only get A while at at -ve max you'll only get B.
I've 2 VCAs and a crossfader I may give it a try.
Not convinced it'll sound any different as AFIACS you're modulating gain with a mix of the upper and lower bound CVs while offsetting the signal with a mix of the 2.
Found the Blasser idea linking the amplitude and pitch together like that quite interesting, quite organic sounding.
peripatitis
My bad I was thinking of CFM's bipolar half wave rectifier.
Btw I feel the bounds thing is probably closer to mannequins poetic descriptions, than substance. But I might be missing something...
windspirit
Nah, its a pretty unique way to make an oscillator. It gives you a
control over the top and bottom point where the oscillator changes direction as well as the slope of the oscillator
Dragonaut
It is just clipping, no? Why not use a uVCA and an attenuated Maths envelope? As the envelope increases in amplitude it allows the oscillator to be clipped by the VCA.
windspirit
It really isn't, best way to understand it is to check out the video:

mbartkow
pugix wrote:
Min/max doesn't change only bounds; it slices off the tops and bottoms of waveforms, changing their shape as well as the bounds.

Neither does rectification, which only removes negative going voltage, setting the lower bound to zero.

A simple VCA can reduce both bounds of any signal.

But what META can do is to take a signal going +/-5V and lower just the upper bound, say, so the signal becomes +4/-5V, or both bounds, so it can swing between -3V and +2V. All the while, if the signal is an audio oscillation, the pitch doesn't change. And the modulations can be other audio frequency waveforms. So here is a type of AM, it seems, that is like an asymmetrical VCA or ring modulator.


If you take a perfect VCA and modulate an audio signal with a CV it will in fact impact both bounds in a symmetrical way. However, you can easily turn it into an asymmetrical modulation by adding/subtracting some % of the CV to the output of your VCA. Hence, this "asymmetrical VCA" only differs from a normal VCA by the presence of the CV leaking to the output. The same applies to audio rate modulation. This may be disappointing, but it does not yield any new timbral qualities beyond normal RM and mixing.
Navs
I find the IFM descriptions willfully difficult to understand, but some of the waveshapes in that video remind me the reversing sync of my Plan B Model 15 and Cyndustries Zeroscillator VCOs.

As you say, Pugix, using a linear crossfader is a great case of reverse thinking. IIRC Rob Hordijk has suggested it for use in his system and it's a method of patching VC slew à la DSG in the Clavia G2 modular. I've also come across it in VC envelope design or Grant Richter's Analog Tracking Generator, where reference voltages are switched or interpolated, rather than the 'driving voltage' itself. The carrier becomes the modulator, right?
pugix
Navs wrote:
As you say, Pugix, using a linear crossfader is a great case of reverse thinking. IIRC Rob Hordijk has suggested it for use in his system and it's a method of patching VC slew à la DSG in the Clavia G2 modular. I've also come across it in VC envelope design or Grant Richter's Analog Tracking Generator, where reference voltages are switched or interpolated, rather than the 'driving voltage' itself. The carrier becomes the modulator, right?


Yes. Normally we think of a VCA as having a signal and a control input, and many of them are designed having in mind an audio signal being enveloped. But if you think of a VCA as a two quadrant multiplier, where one of the inputs has a floor of zero (negative voltage has no effect), then imagine the modulator, the non-negative input, as accepting DC to 20 KHz, and likewise the carrier being anywhere from DC to 20 KHz.

The thing that requires 'inverted thinking' is that the so-called carrier could be just a DC offset. Or in the case of a crossfader, two DC offsets. Fade between them at an audio rate and the frequency of the modulator is transferred to the output, at the bounds set by the two offsets.
smrl
I'm happy to see META sparking some conversation. I helped design it.

Yes, what you're getting by injecting audio into the A and B inputs is AM, though you can do it simultaneously with two signals. Also, by multing the same signal into A and B and using the attenuverter on B you will get 4-quadrant/balanced/ring modulation against the contour generator. You can use the B control dial in the amount of carrier suppression to taste. But none of this is wildly different from traditional AM or 4QM.

One of the ways to use the crossfader is to dynamically control the scale and/or offset of the output, for example if you're using it as a modulator.

Also if you stick audio into the META and it's set to generate envelopes, it'll behave as an envelope+vca for your signal. If you have META set up as an oscillator, and you stick in an envelope, it'll envelope the oscillator. It's just a very flexible circuit, and we tried to create a design that would be very configurable and didn't require a lot of utilities to make it do useful stuff. An obvious case would be if you're driving a CV input with no attenuator, you can scale it with the manual controls. Or use the analog inputs and get "dynamic depth" control. In a way the Via is just an elegant set of utilities with complex control strategies. Richard, you're right -- you can just use an ordinary crossfader and an oscillator or envelope generator and get similar results. It would likely take some fiddling to get the voltages scaled and offset correctly so that it doesn't distort the waveform.

Navs, funny you mention Grant Richter's Analog Tracking Generator. At the beginning we were just calling this project the "interpolator" - the reason there are sample and holds at the input was because we were looking for a way to smoothly transition from sampled values taken from the inputs and be able to chain modules together so they could work in tandem to make multistage envelopes.

I think more of the sonic uniqueness of META comes from the way that the sample and holds are orchestrated at the inputs of the crossfader. Samples are only released when its input's VCA is totally off, so you never get the glitching discontinuities and super bright timbre that you'd usually associate with S+H. You can get decimation effects that are smoothed by drawing a path between the sampled values (and flavored by the wavetables!)
Navs
Got it, thanks thumbs up

Then you could look at Josh Toppobrillo's Mixiplexer as an equivalent too:

http://www.toppobrillo.com/mixiplexer.html
smrl
Yes you could make that equivalence, if you scaled a CV so that it spanned 2 inputs. It's just an ordinary linear crossfader -- but the 2 VCAs that make up the crossfader are independently controlled. So for different Via firmwares we can use it to do different tricks -- we go into more detail on the page below. It's undergoing some revising, in its current state it might not be the clearest explanation but in case you're interested look at the "circuit control schemes" section:

https://starling.space/via/platform-info
joranalogue
Contour 1 allows you to modulate the lower bound in loop mode using the analogue input (top is always +10 V). As the slew rates stay the same, bounds modulation is essentially combined FM and AM; I can't see any way to achieve this using external modules.

You can see it in action on C1 in the demo video, at 33:32:
smrl
Yeah, I saw your design recently and thought it's an awful lot like an analog META! I think we're thinking along the same lines in lots of ways smile

Another note: using 2 VCAs and half-wave rectifiers is definitely not the same, though it's another interesting approach - you're attenuating the positive half and the negative half independently but each has to go to ground so it's like making an extra "breakpoint" at 0V that becomes a new inflection point to your slope. Also each half of the slope would have to stay above and below 0V, with META you can flip the entire waveform around if you want, or have it all positive or all negative, etc.
MUFF WIGGLER Forum Index -> Eurorack Modules Goto page 1, 2  Next [all]
Page 1 of 2
Powered by phpBB © phpBB Group