I'm sure this problem's been around for a very long time in various forms,
but does anyone know the magic combo to get an exact correspondence between
the lens correction coefficients in a pts (panotools/ptstitcher) script
file and Nuke's new-ish lens distortion node?  We are auto generating a pan
and tile dome of nuke cards from the pts.

Without really delving into the maths involved, I have done a bit of
background reading on the lens correction model and the pts script file
format:

http://wiki.panotools.org/Lens_correction_model
http://wiki.panotools.org/PTStitcher
http://www.panotools.org/dersch/barrel/barrel.html

The lens correction coefficients are represented by the o-line options (a b
and c values) in the script and it seems that in all the pts scripts
generated, the b value is the only one which is relevant - the others are
at 0.  But the last URL I pasted suggests that "in most cases" the b
parameter is all that's needed to get "satisfactory results".

What we are currently doing is plugging the b value into the "radial
Distortion 2" knob which is close but looks like it could be a bit
tighter.  We've also tried putting it into the b value of the lens
distortion knobs directly on the card but that looks worse.

Thanks for any insight!

Michael
_______________________________________________
Nuke-users mailing list
[email protected], http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users

Reply via email to