Le_Forgeron <jgr### [at] freefr> wrote:

>
> Natural moves usually try to conserve energy and have no acceleration.
> In such setting, it is unlikely to see the axis of the last rotation be
> moved by the change in the second angle: the gyroscopic effect (and our
> expectation about it) would oppose to such change.
>
The gyroscopic effect: THAT'S an additional real-world 'force' that didn't occur
to me at all-- kind of like the forth (invisible) man in the room  :-O  And it's
quite fundamental to the discussion, IMO. I assume that even a *slowly* rotating
object would exhibit a gyroscopic effect, to some degree. Simple <x,y,z>
rotations in a computer certainly don't take that into account-- which might
explain the apparent difference between using 3 axes of (computer) rotation, and
what we 'expect' to see based on real-world spinning objects in free fall.

For argument's sake, let's say that real-world free-falling rotation of an
object *is* around all 3 axes. Maybe it is, maybe it isn't--but my mental
concept of the effect of the gyroscopic force is that it may 'smooth out' the
jitter/wobble caused by the 3rd-axis rotation-- leading to the expectation that
a computer simulation of TWO-axis rotation looks more 'natural.' Put another
way: If the effect of gyroscopic spin could be included in POV-ray's rotations,
would the apparently weird result of 3-axis rotation 'smooth out' to look more
like two-axis rotation?

Post a reply to this message