If your depth pass is reliably in the same units as your camera, you can
directly calculate the distance from the camera to your axis with an
expression on the focus plane knob, e.g.:
1/sqrt(pow(Axis1.translate.x-Camera1.translate.x,
2)+pow(Axis1.translate.y-Camera1.translate.y,
2)+pow(Axis1.translate.z-Camera1.translate.z, 2))

The above works for ScanlineRender's depth output, with the ZDefocus math
knob set to the default far=0.

If your CG uses absolute depths (really big numbers, background shows
colour of "inf") the above works if you remove the 1/ at the begining and
change the math knob to "depth".

(Technically I think a lot of depth passes are distance to the camera plane
rather than the camera point but the difference wasn't too big a deal last
time I compared)

John Mangia wrote:
> even the best focus pullers in real
> life aren't dead on with their focus

+1!  Removing 90% of the keyframes and leaving some overshoot helps...

-- 
Lewis Saunders
loves to steal axis data from maya rivets
London
_______________________________________________
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