The Official Sequential/DSI Forum

Self-resonant filter tracking query

rb415

Self-resonant filter tracking query
« on: August 28, 2020, 02:03:08 AM »
Just wondering what other Pro 3 owners are experiencing with self-resonant filter tracking. I am getting mine (an SE version, fwiw) to track almost but not quite accurately with both the OTA and Ladder filter.

The situation is a bit complicated to describe - I don't exactly understand what's going on yet, and temperature quite possibly comes into it. But essentially, if I start with the Basic Program init patch, everything turned down in the mixer, resonance on max and filter tracking on 127, the best tuning is when the cutoff is set to a value of 512 (for both the OTA and Ladder).

However, the pitch scaling/tracking is never perfect. This morning, with the Pro 3 powered up for 10 minutes, the 512 cutoff gave me middle A (440Hz) in tune, but with higher notes going progressively sharp until the top C was about 1/4 tone sharp. After another half an hour I ran the osc/filter calibration routine (temperature of 34 reported), and then the lowest note on the keyboard was bang in tune, with notes above it scaled progressively flatter (so the opposite of a 'stretch' tuning) until about 1/4 tone flat.

Interestingly, if I use the octave transposition buttons the low C on the keyboard (assuming that is the best in-tune note) will sometimes be in tune in other octaves. So that suggests to me it's solely the keyboard tracking calibration that is a bit off, not some wider pitch scaling or 'character' of the filters. Perhaps it's possible to scale it a sniff more via the matrix. But my worry is that when calling up patches that use this technique, especially live, you'd never know if you'd get an in-tune result or not.

Adding a little bit of one oscillator encourages the self-oscillation to go in tune and doesn't really change the tone quality, so maybe it's all a moot point. Just curious about it really, especially in comparison to the Moog SUB37 that this Pro 3 is replacing, which has very accurate tracking. (Not that I'd hang on to it for that though - for my purposes the Pro 3 is hugely preferable in just about every other way.)

Thanks a lot! - Robin

Moho

Re: Self-resonant filter tracking query
« Reply #1 on: August 28, 2020, 07:18:07 AM »
IIRC there was a post that said the best results were obtained by using the mod matrix and applying note number to cut off using two slots one with note number to cut off set to +127 and the same again but set to +1

rb415

Re: Self-resonant filter tracking query
« Reply #2 on: August 28, 2020, 07:30:06 AM »
Ah, thanks. Iím annoyed I missed that post - I thought Iíd looked through everything. Iíll try the technique and see how I get on.  :)