Thanks everyone. I’ll chalk up the apparently-inverted envelope amplitude to user error, and re-report as a separate bug if it shows up again and I can reliably replicate.
Regarding the input protection circuit, that sounds plausibly-related. It’s worth noting that both the channel 4 unity and variable outputs go up to the same level (slightly less than +10v) when not plugged into the SSP, and that when plugged in the variable output appears to be clamped at +6v (according to OTool scope), which I suppose is due to the input protection. This kicks in at about +7v (I can make the envelope go up to just over +7v, but any higher causes it to be clamped down to +6v). It’s not clear why Maths behaves funny when the unity output is used instead of the variable output, but it’s probably not worth delving into (I have some other observations I could share about putting other modules between Maths and OTool, such as a 3xMIA, but again, not worthwhile).
In general, the -5v to +5v range seems a bit unfortunate: all of my external envelope generators go above this range:
- Erogenous Tones RADAR is 0-10v
- AJH contour generator is 0-8v (and up to +11v with closely spaced retriggers)
- Maths is 0-10v (channels 1 and 4 can be attenuated, of course, but it’s nice to use the unity out into a VCA and the attenuated output to modulate e.g. a filter cutoff)
I guess this also means that the “standard” 0-8v for pitch is not available. 0-5v is not quite enough to cover the whole range of desired pitches, although it’s probably enough for any one voice (e.g. a bass line won’t span 5 octaves).
Upon reflection, it seems that using external envelopes is the main inconvenience, with the two solutions being:
- attenuate the envelope before sending into the SSP
- instead send external triggers into SSP internal envelopes
The downside of the first is needing attenuator modules, and the downside of the second is not having hand-on wiggleability of the envelope shape (without using additional modules and SSP inputs)