> According to the documentation for POV-Ray's camera:
> https://www.povray.org/documentation/3.7.0/r3_4.html#r3_4_2
> - it seems that POV-Ray performs the look_at transform by first panning and then
> tilting the camera.
> (See 3.4.2.1.1 Location and Look_At)

I had read this, but I did not draw the necessary consequences.

> The code that I posted did this in a single reorientation.
>
> If this reorientation is split up in two steps (first panning and then tilting),
> the results seem to be the same for the created user defined camera and the
> built in perspective camera.
>
> You can do this by changing the expression for the LookAtTransform in Ingo's
> ucdLookAtTransform macro, like this:
>
>     #local LookAtTransform =
>         transform {
>             Reorient_Trans(z, vLookAt*(x + z))  // Pan
>             Reorient_Trans(vLookAt*(x + z), vLookAt)  // Tilt
>         }

Thank you Tor for this fix. Now the user_defined camera transform is working
properly!

igmar

Post a reply to this message