|
|
Le_Forgeron <jgr### [at] freefr> wrote:
> Is there a need for deep control or can a simpler version be enough ?
>
> 1. can origin of the axis always be <0,0,0> ?
> Or should it be customisable ?
I think for most things applying the warp and then a translate would be fine.
> 2. can the direction of the axis be specified ?
> Or should it always something like +z ?
Again, I think for most things applying the warp and then a rotate would be
fine.
> 3. can the handedness of the rotation be always the same ?
> Or should it be customisable ?
Why not just have a signed amount?
> 4. can the distance between two complete rotations be always the same ?
> Or should it be customisable ?
Indeed, I thought of this, and perhaps there's a whole new idea that may dwarf
even this small level of complexity. *
> 5. should the rotation always occurs ?
> Or should it occurs only for a bounded range along the axis ?
For now, I'd say keep it simple, unless the choice of a bounded is not much more
complicated.
> 6. should the rotation be of constant speed along the axis ?
> Or should it use something like *dist_exp* from
> http://wiki.povray.org/content/Reference:Warp#Mapping_using_warps
> with something else than 1 as its value ?
An excellent question, and one that almost hits on my thoughts on #4 (*)
Perhaps to allow added user-control, with user-defined functions, there can be a
warp_map, such that the affected regions and degree of effect of a warp can be
controlled by a spline.
warp {
rotate x linear // quadratic cubic bezier
warp_map {
[0.00 pi/2] // per POV unit
[0.50 2*pi]
[0.75 3*pi/2]
[0.00 pi/2]
}
}
This may be way too complicated, as there are a lot of moving parts under the
hood, and I don't know how much this would slow things down if applied.
Post a reply to this message
|
|