A joy to listen 
1 Like
Tonight’s fun involved using Surge XT TreeMonster to extract pitch and envelope from an input, then feed those as CV signals into Sapphire Tube Unit. Stir in some slow chaos and mix together to obtain:
2 Likes
Some drone with bass string plucks and bells:
6 Likes
I gotta start using more of those chaos generators.
1 Like
Tonight I decided to try creating a complex synth voice using multiple voices and slow chaotic modulation. Then I performed the patch with an Impromptu Twelve-Key using the same V/OCT for all the voices.
5 Likes
This was an experimental patch to verify behavior of my soon-to-be released new module Sapphire Env. It turned out to be more interesting than I thought. Env is a pitch and envelope detector, and here I use it to play Venom VCO Unit to mimic an Elastika that is tuned to sound something like a sitar. I use linear FM modulation on the VCO Unit at audio rate from a post-reverb signal from Elastika.
Another fun quirk of this patch is the use of Chaops to help two chaos generators (Frolic and Glee) dramatically modify each other’s behavior using the MORPH, SPEED, and CHAOS controls.
8 Likes
Wonderful sounds and textures!
1 Like
These sounds are magnificent! Almost sounds like a steelpan .
1 Like
First of all:
Very much a fan of all sorts of waveguide/resonator based stuff…like you have given us ( Elastika, Tube Unit). So much potential still pretty much untapped there. The success of the various Mutable Instruments resonators have also shown us a lot of the potential and the demand/appreciation of resonators. Some resonating filter based resonators are also great.
Just a flexible 2D Waveguide mesh/matrix (1 x 1, 2 x 2, 3 x 3…) with inputs and output at ends and junctions and control over basic variables like lengths, stiffness, dampening (and/of filtering), polarity would probably open up quite a playground.
Using VCV Library - Sckitam WaveguideDelay for this is not ideal. And not just because of the one sample delay introduced by each cable. Not-to-complex complex and pretty usefull and great sounding Waveguide/Delay based resonators with various excite/pickup/feedback points can be created with WDelay (e.g. strings/tubes). But the all modules and cables needed for more complex Waveguide/Delay network experiments can get confusing real soon.
But I digress…
Using a some complex audio signal as a modulation source for LFM is fun. Either an ‘in tune’ signal, e.g. feeding back from further downstream in the signal path or modulating from some unrelated audiosignal.
FM/PM is pretty susceptible to exact tuning and integer ratios. And soon as you really step away from integer ratios, the inharmonics appear. Also, due to the non-linear nature of FM, high amplitude overtones/peaks can appear in ‘unusual’ places and move around in unusual ways (generally simultaneously traveling up and down the spectrum). Even when using integer ratios, but with larger distances between them then +/- (0), 1, 2, 3, 4…
Note that sidebands are created for ALL combinations of harmonics/frequencies in the spectrum of both ‘operators’ / signals involved. So…huge amounts are generated when using complex spectra, even more at higher modulation amounts.
All this is also usefull for percussive and/or near chaotic transients.This kind of percussive wide spectrum attacks transitioning towards by less but inharmonic and/or unusual spaced overtones can often be perceived as metallic.
Side note: my old Yamaha SY77 (but also the SY99, that could even load samples) could use the AWM (sample based subtractive engine) as a modulation source for the AFM engine (phase modulation). Pretty unique feature…especially in those days (1989).
2 Likes
I have this feeling that creating mutually interactive feedback loops brings a patch alive, even when you do it at LFO rates like with the two Chaops. I will definitely try this. I didn’t have any ideas for Chaops before; now I do. This is just the kind of thing I like.
1 Like
I have that feeling too! It’s like the patch is a creature that has its own emergent personality, not something I planned ahead of time.
1 Like
With feedback loops things can soon spiral out of control. Be it audio signals or control signals.
Generally the challenge is often about managing (min/max) extremes, and about predictability versus complete chaos.
If all else fails, simply introduce hard/soft knee compressor/limiters in the signal path. Or, to achieve similar effects, use waveshapers/limiters like Instruō tanh[3] or Bark Clamp or any other amplitude remapper to enforce limits.
By their nature these amplitude controlling/remapping tools obviously are able to change the waveshape of the signal and thus change the spectrum. This of course depends a lot on the shape, threshold/knee, amount and timing of the amplitude modulation/waveshaping applied in combination with the amplitudes of the input signals. When applied to audio/audiorate signals, these tools/techniques might lead to (very) audible results. But…distortion effects are not necessarily a bad thing.
Also, (lowpas/bandpass/highpass) filtering can help in limiting unwanted chaos in audiorate modulations scenario’s. Filtering can help to determine/limit where and how much the resulting spectrum is affected. Filtering can also reduce amplitude (depending on filter resonance/feedback).
.
I’m always trying to tune audio feedback to the root note. I’ve tried all kinds of things in the feedback path. Sometimes I can tune it by attenuating the feedback to just the right amount. Always a struggle though.
IO went and tried to download the builds but they say expired.
The Sapphire Env module is now officially released on the VCV Rack Library, so you can get it here:
4 Likes
I had fun with this deeply layered, slow-chaos modulated drone chorus. This is definitely my happy place!
Here is the VCV Rack patch.
5 Likes
I’m having fun with Sapphire Echo way past my bedtime:
The patch is here if you want to play with it:
6 Likes
Lovely patch. Would love to have this running in the 4ms Meta. Any chance of a fixed structure version (2,3 or 4 channels) targeted specifically for the Meta (AFAIK it won’t allow adjacent module linking but I expect you know that!)?
Dan and I collaborated on a high level design for how to implement expanders on the meta module, but I don’t think Dan has a timeline for implementation yet. So that would be a tricky call for Don to make a major change to the code, or to wait for Dan to implement expanders.
In addition to having to rework the expander chain into a single module like Dave correctly pointed out, I think another challenge will be CPU and memory overhead. On my Linux desktop system, each Echo tap uses about 2.8% CPU. The MetaModule hardware has much less number-crunching power than a full-blown desktop system constantly blowing 100 watts of warm air out the back.
I know from helping port Elastika to MetaModule that this amount of CPU overhead can saturate the MM’s processor and cause stuttering.
Also, Echo uses a significant amount of memory. Each tap has to allocate enough memory for 16 channels of 10 seconds each. At 48 kHz sampling rate this works out to roughly 30 MB of memory per tap. I’m not sure how much memory is available in MetaModule, but this would have to be evaluated.
So, it’s an interesting idea but my gut is telling me it would be a lot more work than I can commit to anytime soon.