I have tried many many different joystick utilities, Pinnacle comes very close to what I'm looking for. I tried a demo of it last year and again today, and I'm encouraged by the progress, but so far it still doesnt do anything better than my other software does in terms of controlling the game. Here are some suggestions:
Pinnacle has finally added shifted states for joystick axes, BUT you dont have a seperate sensitivity for each state in it. For instance if you hold down the shift button then your joystick could become more sensitive, let it go and it returns to normal. [That way you can have a fairly crude but effective aiming method for playing shooter games.]
Another feature that we are dying to see is a user editable response curve for the axes, I have seen this in only a very few high-end or specialty utilities but its really useful. For instance you could adjust for your stick to be low sensitivity all the way until you reach 90% of its range where it becomes highly sensitive because of a steep response slope. This helps aiming dramatically in shooter games, they become as playable as with an actual mouse since you can slowly aim using most of the range in the stick but also use it to turn quickly when needed by shoving the stick all the way to the edge, or otherwise tailoring it exactly how you want it to behave. Another example is that most joysticks are really stiff in the deadzone area to prevent them drifting, editing the response curve can make it seem less stiff and "poppy" when moving away from center and still give you the proper sensitivity throughout the rest of the stick range.
edited here, I was going to suggest force feedback support but I see now it's already in there![]()





it would be closer to what I'm after to have it more like 1% or even a fractional percent of the range at either end, I would want it to use the least possible amount of range at the edge. After trying this I got an idea to make a banded "axis command" with each zone having a sensitivity modifier but pinnacle forces your zone sizes to be somewhat large. This might be a good way to fake up a response curve if the zone sizes could be made smaller.
