Circular expo for acro and altitude #75
Conversation
…es before applying expo. This prevents extreme control inputs in the corners, and gives the pilot a round input space which makes sense. Make expo not leaner + ^3 but rather linear + ^5 which should make center stick much more linear.
|
Thank you for submitting this merge request. I read through the code but did not test fly. I have spoken to Paul about this change as well. I think the idea to apply the expo on the vector length instead of the individual components is good. I don't really like the hard-coded change to the expo function to make it x^5 instead of x^3. This changes the response of the expo in all places expo is used. The pilot feel for an expo value with this change would be different in this branch than any other branch of PX4. If any user has an intuitive feel for what a certain expo value means that mapping would be changed. Also a user couldn't go back to the old curve (x^3) without a code change. Can you find an expo parameter for the existing x^3 curve that is close to the response you want? If not what do you think about making a PX4 parameter to choose between the x^3 curve and x^5 curve? |
Fix expo function to maintain vector angle & correct magnitude in circular geometry (instead of calculating components in cartesian). This prevents extreme control inputs in the corners, and gives the pilot a round input space which makes sense. Make expo not leaner + ^3 but rather linear + ^5 which should make center stick much more linear.