One idea could be the following: Take the analog noise (A output on the NOISE module), run it though the ATTenuator and from there to the LOGIC module; here, use the XOR output. The LOGIC module works like a threshold here; with a certain signal level (determined by the attenuator setting) more or less "peaks" of the noise signal get over the threshold and so you get a full 5V trigger signal at the XOR output. In theory so far... let us know if it works like this!
With one attenuator it was very difficult to find the spot where it's not crazy fast. It was super fast or no movement at all. I sent it through the second attenuator additionaly and it was a bit easier, still crazy fast or nothing but the tiny window of usable speed was a bit wider and a bit easier to set, but the knobs were VERY sensitive for the slightest touch. Such signal was sent to BEAT DIVIDER to make slower variants if necesery. So I think it works, would have to compare the results with the behaviour of CR output to determine the differences, but the main goal to have random triggers seems to be achieved. Thank You very much robertlanger !
interesting findings... One additional idea could be filtering the A noise; this might reduce the rate of the random pulses also. With an additional divider please keep in mind, that with higher division rates you get an average over time, making the signal the more a regular squarewave the higher the division factor is - I learnt this the hard way while writing the code for the CR out signal; had the same approach first ;-) Amazing what you can think about random pulses, isn't it?