Just wrote this over at Patreon in reply to your issue with trigger voltages, but copying it here as well for reference:
"The trigger pulse depends on the input voltage; the next stage compares the filtered signal with a 2.1 V threshold, so if that 10 nF capacitor doesn’t charge up enough, the signal will spend very little time above that threshold, which affects the decay/release cycle as well.
"To make this really robust you probably need an input buffer to make sure you have a constant trigger voltage no matter what you connect it to, but you could try messing around with the threshold voltage divider (the 10k and 47k to the left of the opamps). E.g. try soldering in another 10k in parallel with the existing one; that would lower the threshold to a bit over 1 V.
“Another option is to make the threshold adjustable, by replacing the 10k/47k pair with e.g. a 100k trimmer between 0 and +12 V, with the wiper connected to the negative opamp input (you can use the same trimmer for both opamps).”