Why not using splines?



news:web.40a4ebc89de6780ab2365a790@news.povray.org...
> I actually work on a bigger project, showing a gothic cathedral. I am
> planning a cam-flight from far away toward and round the building, later a
> walkthrough inside the building (90-120 secs).
> For animating the cam-track, I was not very happy whith the implemented
> POV-animating possibilties.
> I used POV-Ray many years and even under DOS I did a lot of scene scripting
> by using external script-generating tools. First POV-vrersions had no
> #macro, so I scripted some tools, generating from a text file whith mixed
> POV- and Expression-code complete POV-scene files. Since I am using
> UNIX-System my possibilities got a lot better, especially using perl:-)
> Back to animation: to creat my cam-track, i make a simple text file, where
> all relevant cam-options are listed in keyframes e.g. like this:
>
> frame distance<0,0,0>   height  rotation  look_at_x  look_at_y look_at_z
> 0000        500           200      0          0           23       0
> 0200        200           100      45         0           23       0
>
> For my current project for each keyframe I declare about 26 options
> including cam-angle(for Zooming and wide-angle), focalblur, fog, lights as
> dynamical options.
>
> This KEY.INC-file runs through a perl-process which generates a
> POV-include-file for each frame of the animation by interpolating curves
> through the given control points in the keyframes. After the
> POV-Include-file is generated for a frame (or a set of frames), these
> frame(s) are rendered and attached to the animation. All these processes
> are managed by one perl script and can be stopped, continued or changes can
> be made to the cam-track whithout losing all rendered data.
> At the moment, this wroks only with global keyframes, but I am working on
> the implementation of the possibility to key each single option on
> different frames.
> Also I'm working on the possibility to change the orientation of the cam
> between global look_at-mode and relative rotate_xyz-mode to the actual
> moving vector of the cam.
>
> Anyone tried this or something similar?
> Coments welcome ...
> Siggi
>
>
>

Post a reply to this message