Could you fake the ramping/realtime feedback, and then periodically have S+ evaluate the real analog feedback? You'd then take the real (scaled) feedback and update the interface and the value that would be ramped from next. That would keep S+ from trying to deal with rapid ramping calculations. So instead of a CHANGE event constantly monitoring the analog input to S+, there'd be a separate evaluate_now input that you'd pulse from an OSC or whatever.
?
--
Josh Winn
The LiquidPixel Group
josh@LiquidPixel