With more thanks and respect that can possibly be typed on a QWERTY keyboard, please consider the following feature request for Karma FX Synth.
I would like to see more asymmetrical clipping options on the Distortion Effect module - similar to but expanding upon Synthesizer.COM's Q130 Clipper http://www.synthesizers.com/q130.html.
The current KarmaFX implementation of Hard Clipping produces an asymmetrical signal with the absolute negative levels less than the positive levels. The result is the addition of the higher degree partials and thin buzziness. To tame the buzz, I end up applying a 2-pole filter, but that compromises the use of the distortion in the first place.
The Synthesizer.COM module produces asymmetrical distortion by clipping EITHER the positive or negative portions of the signal. The result is less buzziness with the ability to boost the lower degree partials by overdriving and no need to apply filtering to tame the higher partials.
Now imagine that you can apply hard OR SOFT clipping to the positive OR negative (or both) portions of the signal. Is there ANYTHING out there that does this now? I'm not sure, but I'd love to have it inside KarmaFX Synth.
Thanks for the suggestion!
That's a really interesting feature-request!
So would it be sufficient with a "clip switch" drop down /
with a "+", "-", or "both" clipping option in the distortion module ?
I need to see how this would fit with the different distortion modes...but otherwise it wouldn't be that hard to do.
(I'm revisiting the KarmaFX board after a week's hiatus.)
I'm thrilled that you like the positive/negative/both polarity distortion idea. The Metasonix R-51 does something similar (with real tubes), except (I believe) that it only clips the positive portion of the wave to produce the asymmetry. The Q130 offers more options but no tubes.
I'm using KarmaFX Synth's existing Hard (Clipping) for waveshaping (pre-filter) with very satisfactory results. Yet I still find myself wanting more control.
I'm thinking that, rather than add the polarity options to the Effects/Distortion module, the introduction of a new Waveshaper module may be more desirable. In this way, users who are expecting straight-ahead distortion may get their effect without any additional fuss.
The Waveshaper module could be applied to audio and control signals, and contain a host of non-linear algorithms which could grow over time. When applied to audio signals, the module becomes a "timbre modulator" similar to Wiards's "Waveform City" module. "Hardclip Positive," "Hardclip Negative," and "Hardclip Both" would be three options/dropdowns/algorithms available on the new KarmaFX Synth Waveshaper module. (No need for saparate LED's.)
The new module may be a good way to breath new timbres into the Oscilators' trusty waveforms.
If you like the idea, I'll go a bit further on how the "Drive" parameter on the current Distortion module can be replaced by a "Level" parameter on the Waveshaper.
Thanks for listening,
And I like the idea. A new module does take longer to implement though.
If a new module is to be written should it perhaps be a more general waveshaping module, i.e., I'm thinking where you edit/draw your own waveshapes ?
Yes, I'm interested in your take on Drive vs. Level.
When I overdrive a circuit, I need to boost the energy of original signal to bring it up to the level where the clipping starts. I get almost immediately satisfying results this way, BUT, what I've found is that I've fooled myself. What I've really done is to increase the LOUDNESS of the signal, which I confuse with the sweetness added by the clipping. When I compare apples-to-apples, or in this case, loudness-to-loudness (reduce the distorted level to the same level as the non-distorted tracks), I am less satisfied with the result. This means that I'm tinkering with Drive in the Distortion Module vs (Loudness) Level in the module's output or out in the mixer. This is not the way I want to work.
Instead, I want to bring the clipping/waveshaper Level/Threshold down to the signal (rather than vice versa in the first example). I haven't really changed the loudness of the original signal, but I have redistributed the energy of the (sweetened) signal in a way that I can hear without confusion. I can then bring that exact signal up into the mix where I want it.
This is the "control" of the timbre modulation that I'm seeking.
Z3ta+ has a "Shaper" window where you can apply a good number of algorithms with different intensities to "Vintage" waveforms and watch them become mangled before your eyes while playing the resulting sounds on the QWERTY. This is a very instructive exercise in distortion, and it takes place pre-filter - exactly where I want it.
BUT, the distorted waveform is STATIC; the clipping (or whatever) cannot be modulated programatically (despite Z3ta's impressive modulation matrix). This jives with Z3ta's characteristic, synthetic sound, but is not well suited to the realism that I seek. This limitation can be circumvented effectively by using two or more Waveshaped Oscillators, and varying the levels between them with an EG. This is a good amount of work however, which likley can be accomplished more easily (and with less CPU) with a Waveshaper module against a single Oscillator with a Level/Threshold that is modulatable.
EDIT: The Waveshaping level/waveform can controlled via CC.
Drawing the transfer function is not as big a deal for me as watching the time-domain waveform get warped asymmetrically (or perhaps admiring the changes in the frequency-domain partials). I can see where some users may want to draw (and save) their own transfer functions in a window with a mouse. That could be a nice feature, but I would be more interested in a rich collection of algorithms from which to select.