Alright lets fire this one of with the rolling ball exercise. Perhaps not the most exciting content for a polybook but it has to start somewhere. :)
I found this one pretty tough and I am still not completely convinced by my final results. I wanted to refrain from calculating the correct rotation value for the distance covered and do it by feel. In hindsight I really don't know if that was beneficial to the point of this exercise or actually one of the things that made it unnecessarily hard.
What helped me in the end were a few different things:
- Starting with linear interpolation to get the initial rotation value in the right ballpark
- Setting viewport display to "render only" so the rig does not distract you (this one was huge for me)
- Returning the units for rotation to degrees once the keyframes were in place to make the fine tuning a lot more manageable.
- Having an accountabilybuddy
Hope it helps someone although I feel these might come down to preference more than anything. :)