I got my 18 HP Permutation last week (thanks for the fast delivery Mattech Modular!). One of the reasons I swapped my OG MTM TM for this one was to have the Write & Clear switches under CV control. I thought it would be fun to "program" it with a gate sequencer (René Mk2 in my case). I wanted it to play a specific (gate) sequence and then let it gently drift away, only to be pulled back into the original sequence again. After a bit of headscratching I finally got to a working solution. It works as intended, quite fun. Since you can actually program the gates, you can effectively program the note sequence at the SEQ output. Since that is a result of summing the bits, if the bits are in the same place, the sequence will be the same. So that gives a bit more predictabilty and even allows you to return to a sequence you like if you take note of the gate pattern. Maybe this could be of use to people who want to play Permutation live but want a bit more control.
I've expanded a bit a bit on my original idea and added some sort of preset states to go to with Pressure Points and a modDemix.
Patch description:
Maths #1
Ch.4 is the main clock. The EOC output is patched to CH.1 In and René X-Channel Clock In.
CH.1 is the clock for Permutation. It is delayed a bit by turning Rise a bit from CCW and then taking the EOR output to Permutations Clock Input. The delayed clock is essential for the functionality; if the clock of Permutation is in sync with the Write or Clear actions it won't work. I've taken the same EOR output also to a Tempi and clock all other stuff from there (like percussion) so that it is in sync with Permutation.
René
I've programmed the gate sequence as marked by the white dots as an example, but it can of course be programmed to whatever pattern you like.
The Gate output from the X-Channel is patched to Maths #2 Ch.1 input (for the Clear functionality) and to modDemix Ch.1 (for the Write functionality).
I've used all 16 positions on René and set Permutation to use 16 Bits.
Maths #2
The Gate signal that arrives at CH.1 input is inverted by turning the attenueverter of Ch.1 fully CCW.
Ch.3 provides a +5V offset by turning the attenueverter fully clockwise. Ch.1 and Ch.3 are summed together at the SUM output and this is patched to modDemix Ch.2.
Ch.2 and Ch.4 have dummy cables at their respective outputs so they don't go to SUM.
Pressure Points
PP is used as a preset states switch. I set up Stage 1 as the "default state". It sends +8V (top row) to the modDemix Ch.1 Strength input (Write functionality) and +5V to Ch.2 Strength input (Clear functionality). The middle Channel is patched directly to the Permutation Shift input. It is turned fully CW so it provides a +5V offset to the Shift parameter on Permutation. The Shift knob on Permutation is set fully CCW. This way, the PP knob on the middle row functions the same as the knob on Permutation, i.e. fully CCW is [CYC] and Fully CW is [LIN]. In the middle is completely random. For the "default state" I set this fully CW because I don't want randomness there.
I've set other Stages with different settings. For instance, Stage 2 is Write and Clear both off (CCW) and the middle channel just a tad turned upwards from CW so that introduces a bit of randomness.
When I want to go back to my original gate sequence I press the Stage 1 touchplate.
modDemix
Ch.1 output is patched to the Permutation Write input.
Ch.2 output is patched to the Permutation Clear input.
Both VCA's are being opened/closed from Pressure Points (Top row to Write, bottom row to Clear)
Summary
Although it works as intented I find that the Shift parameter goes rather fast from no changes (CW) to a lot of changes (turned a bit upwards from CW). With my old TM it only started to change with the knob set around 02:00 (noon being completely random). Permutations already starts changing before I get to 03:00. So there seems to be a smaller "sweet spot".
I also had Permutation going haywire (go full random even with Shift turned CW) and not responding to any of my Clear/Write signals at their respective inputs. The only way to stop it was by manually pushing the clear switch. But that only happened once so I might have done something wrong there.
This is the patch on ModularGrid. I've started uploading some "Outline" versions of the modules there so I could make my own patch sheets easier:
