I have no idea why this happens. I’ve left the patch as is for now, and will try to get deeper into what’s going on, but it was interesting enough to warrant mentioning it here in case someone has a better grasp of why a filtered version of the oscillator (or to be more precise, an output of a filter used as a wavefolded oscillator) would be a good thing to plug in the frequency input of a comb filter.
Yes of course, I’d be happy to share! I’m using the ADDAC215 extensively for such uses when I patch bowed string sounds, but it can be used for any similar purpose. Just to clarify, does your question revolve around how the ADDAC215 should be set up for such a purpose, or how it would be seated within such a patch to serve as a RoC module? I’ll try to answer both separately but let me know if you need me to expand on either direction.
ADDAC’s 215 settings, inputs, outputs.
A clock source is necessary to get the ADDAC215 process going. That needs to be patched in the Trigger A input for this method*, and the operation needs to be set as Trigger A alternate 1 > 2. The module should be set as a dual T&H, with your CV source (more on that in the next paragraph) plugged at Input A. You’ll get the RoC of your CV source at the Difference output. You’ll most definitely need to have some slew engaged at the B section to tame the output’s changes. The speed of the clock and the amount of slew is up to taste.
*Notice that I said “for this method”. You can get similar results with S&H instead of T&H and a common (re: non-alternating) clock at both sections. It’s a different approach. You’ll need to slew section B again but this time the purpose is to delay the timing the calculation happens, where in the above patch that’s determined by your clock speed only.
The basis of the RoC patch
I usually use a slider module, namely the Ladik U-081 (disclaimer: worked with Ladik on this design), or a Soundmachines LS1, as a bow. I use that as the CV source that goes into the ADDAC215. I also mult the slider’s output into the Ladik Derivator module to get gates respective of whether I move the slider upwards or downwards. I will plug the ADDAC215’s Difference to my VCA and probably the filter of my sound. The speed of my gesture will determine how quickly I bow the string. If applicable, I’ll also repurpose the Average output of the ADDAC215 to define the position of the bow on the string.
Extras
The Derivator’s outputs can go to lots of places: how about changing the value of a Slew module applied on the pitch values to have different slew for each direction? I sometimes add part of a gate in with the pitch value to get a jump on the note depending on the bow direction.
Generally speaking how you patch the components is up to you. I try to read up on actual techniques and try to replicate them. As I said earlier, for me the realism of a physical modeling patch is more than 50% based on the way the patch is played. A shitty sound with proper envelopes will sound far more real than an amazing sound with rigid ADSR.
Anyway, I hope the above is of some service. Please let me know if you need more details or if something doesn’t make sense.
