The scaling on bipolar modulations should use the 0-center as axis, applying the selected scaling curve to positive values and flip-mirror(!) its behavior to negative values. The current behavior is not logical, except for the linear curve (which automatically provides the suggested behavior by its linearity).
Makes sense, but such a change would affect all existing meta device scaling too e.g. controlling track pan with a Hydra, using non linear scaling.
I think that’d be fine. ^^ I wonder if anyone expects a scaling curve on bipolar values to be spread the way it is now. I mean, I really(!) wonder. Imo fixing this would just correct the behavior to a state that actually everyone would expect anyway. Wouldn’t it?
Some other thoughts, anyone? What do you guys think?
Just as a side note: For 4.0 please drop the downward compatibility priority. Everyone wanting to use his old files could still use an old Renoise version. One hard cut could make a lot of things way easier.