Le_Forgeron <jgr### [at] freefr> wrote:
> Le 12/03/2018 à 17:58, Bald Eagle a écrit :

> For Povray, either
> * you compute the shear matrix to project the red box on to the road.

I'll have to think about that a bit, but I believe I could work that out - I
hadn't thought about shearing a text object.  Nice suggestion  :)

> * or you use a 2D pattern in plane of the red box directly on the road
> (and let the software do the work)

Ah yes, having recently played with labeling the values of the crackle pattern,
I should have thought of applying that same technique.
A "2D" pattern that stretches infinitely "deep" should intersect the plane of
the road and give approximately the desired effect.

> ... it might not be worth to
> compensate for the horizontal distortion.

Probably not.

I have actually seen a fair number of rounded and lowercase letters, as well as
stuff like pedestrian and wheelchair glyphs.


Though I'll have to crunch some numbers to see what the actual difference in
scaling between the top and bottom is.  I made essentially the same diagram that
you did, with the same red plane perpendicular to the camera_location-to-look_at
vector, so I could see what sort of right triangles I'd be working with.
Comparing the projection of the top section with the projection of the bottom
section is what suggested that the difference in scaling might be significant.

Then I had to get back to work in RL...   :|

I suppose if I actually _projected_ the text onto the pavement, or at least
declared such an object that I could then use as an image map...
I'll have to look back over Kenneth's work on the cubic pattern...

Post a reply to this message