Cella Announcements Thread: latest version 2.0.5

[Cella (Victor Kashirin)]

Hi there, I’d like to introduce my small collection of modules released under name Cella. Current release is 2.0.1.

Bezier is a smooth random voltage generator based on Bezier curves. Heavily inspired by Random Bezier Waves / ADDAC507 module for Eurorack, which ADDAC designed in collaboration with Rijnder Kamerbeek aka Monotrail. Module consists of one curve generator with adjustable slope - from rounded to spiky. New random value is sampled at set frequency, and module extrapolates to it from current value following curve of desired properties. Module can also sample external signal, thus working as a sort of slew limiter.

Euler is a simple module that calculates rate of change of incoming signal. Specifically, it calculates angle of a slope of incoming signal, normalises it by 90 degrees and scales to -10…10V. FREQ parameter allows to adjust rise (volts) and step (seconds) to a meaningful ratio, related to general frequency of incoming signal. Module can be used to derive related control voltage, and might be fun to use in some cybernetic patching.

Modules are shared via VCV Library.

Source code is available at GitHub.

And the manual is here.

I have a few more modules on my mind, so updates to the plugin will be announced here. This thread can also be used for submitting any questions or issues you have found. Issues on GitHub are also very welcome.

[rsmus7 (Stephan)]

thanks a lot for these, and welcome to the vcv community

[Cella (Victor Kashirin)]

Happy to announce new 2.0.2 release of Cella plugin and addition of a new module that I’m very proud of - Rich (in the name of Buddy Rich, master of dynamics).

Rich is an expressive AD envelope generator with stepped or modulated accent. Step parameter specifies how many increments accent will have if it is triggered continuously. Positive number of steps will provide expected loud accent, while negative value will make an inverted accent that subtracts from baseline envelope level. Stepped accent levels can incrementally rise or fall, which is controlled by the button at the center of module or a gate to the input below it. If envelope is triggered and accent input equals 0V, accent value is reset.

Here’s basic visualisation of output when accent is triggered:

Importantly, Accent input accepts not only triggers, but any CV signal between 0V and 10V, and further modulates accent level, allowing for any desired accent pattern. You can use Accent input for modulating velocity of the whole envelope, simply by setting baseline to 0.

Port ACC outputs current value of the accent, but independent from ACCENT level parameter. Hence, you can set accent level to 0, and still use accent value to modulate external or internal parameters, for example decay length of Rich.

One quite helpful and unique usability feature of the module is its tolerance for delay of few samples between envelope and accent gates caused by different length of cable chains - even 1 sample delay is usually enough to cause other modules to miss the accent. Rich will register accent trigger even if it arrives up to 5 (or 10, it’s configurable) samples before or after the envelope has been triggered. Thus, you don’t have to calculate exact delay in samples and add mult’s or helper modules to ensure that accent and envelope register together. And of course, you can switch this feature off.

Here’s a demo of module’s capabilities. Jump to 1:10 see accent explanation:

Rich - Envelope Generator With Natural Accents for VCV Rack

I find this module very musical, and it seamlessly adds dynamics to any sequence, making it more alive. I’ve wanted to have such module in VCV for a long time, so here it is. I hope you’ll enjoy it.

It’s already available in the library: Cella - Rich

[rsmus7 (Stephan)]

Rich is a very usefull module and you made a very nice tutorial!

[Cella (Victor Kashirin)]

Happy to announce new 2.0.3 release of Cella plugin and addition of a new module - Twin Peaks.

Twin Peaks in VCV Library

Twin Peaks is a dual resonant filter inspired by the design of Rob Hordijk’s TwinPeak filter and Klangbauköln’s Twin Peak Resonator. The design is straightforward: the same input signal is passed through two resonant low-pass filters, A and B, and their outputs are subtracted, producing a band-pass response with two peaks. This design allows the filters to pass each other, meaning it doesn’t matter which filter has the lower or higher cutoff frequency — the signal still passes through.

It’s ideal for pinging, and here’s a recipe:

1. Feed it stepped CV from S&H, Random module by VCV, or Rungler output from Benjolin Oscillator.
2. Set resonance to ~78%, adjust cut-off frequencies and increase Track A and Track B parameters a bit.

It’s a great companion to Benjolin Oscillator by Venom, and two of them practically make a Blippoo Box.

Here’s a short tutorial & demo:

Twin Peaks by Cella - Dual resonant filter for VCV Rack inspired by Rob Hordijk's design

Under the hood, the module uses the DSP core from Audible Instruments’ Liquid Filter but adds the option to use an 18dB 3-pole filter output, which was used by Rob Hordijk in his TwinPeak filter.

Also, in this release dark panels were added. Unfortunately you can’t choose between light and dark theme via context menu yet - that’s coming up with the next release.

Hope you’ll enjoy the new module!

[DaveVenom (Dave Benham)]

Sweeeeet!

Benjolin Oscillator and Twin Peaks - a wonderful pairing!

[codyge]

These really look lovely! Can’t wait to take them for a spin, and interested to see what you come up with next.

[fulhjort]

Added Twin Peaks to a patch I started this evening, it really made it pop (as much as harsh noise walls can pop ;)).

[Cella (Victor Kashirin)]

Thanks Cody, your plugin is a big inspiration for creativity!

[Cella (Victor Kashirin)]

Haha, glad it’s been of help! It can definitely get crazy.

[Cella (Victor Kashirin)]

Happy to announce new 2.0.5 release of Cella plugin and addition of a new module - Resonators.

Resonators in VCV Library

Resonators is a module that features four pitched resonators based on the Karplus-Strong algorithm. It is designed to create rich, resonant sounds by simulating the behavior of plucked strings. Functionality is inspired by audio effect of the same name found in Ableton Live.

It’s not polyphonic, but has multichannel output WET that provides signal from each resonator for further manipulation such as panning. Also multichannel input of FREQ I can be used to tune all four resonators with one cable.

Here’s a short tutorial & demo:

Resonators by Cella - Quad resonator for VCV Rack

Also, with 2.0.5 all documentation moved from PDF file to Github for easier access: manual

P.S. 2.0.4 was a bugfix release - it fixed issues with Twin Peaks module that could occur when # threads > 1.

[fractalgee]

Twin Peaks is seriously amazing, gotta see what Resonators can do… Love me some resonators after all…

[cosinekitty (Don Cross)]

Resonators sounds very cool. I’m going to experiment with it in my own patches.

I took a look at the source code, and if I understand correctly, you are using delay lines whose length is tuned as close as possible to the desired resonant frequency, but then also using linear interpolation between samples to correct to the exact frequency. Is this right?

Does it help to have a prefilter (like you use Twin Peaks in this video) to reduce distortion at the higher frequencies due to linear interpolation? Or did I miss the point entirely while reading your code? I’m genuinely curious because I want to up my game in DSP theory as a fellow VCV Rack plugin developer.

[DaveVenom (Dave Benham)]

Sounds like a good separate topic for the development category

[Squinky]

but, yes, fraction delay and being out of tune is a big topic. lots written on that. I think there are several techniques that are more in tune than linear interpolation. that said, I don’t know how badly LI is out of tune at high frequencies…

[Cella (Victor Kashirin)]

You are right, linear interpolation is used to obtain fractional delay for finer tuning. More advanced approach would be to use Lagrange interpolation, but I went easy way for now.

Mind you, I’m not a DSP expert, just learning as I go.

Twin Peaks works as a sound source in the demo rather than a filter, so that wasn’t the purpose.

[Squinky]

I think IIR filters with an allpass response are sometimes used, and various approximations of a perfect FIR filter…

[LarsBjerregaard]

That’s nice work!