0
$\begingroup$

Problem

I've been reconstructing point clouds from blender's Z-pass outputs; saving to .exr to capture the full detail. I've noticed that when I use the EEVEE engine flat surfaces are reconstructed flat - but using cycles the same captured Z pass includes radial distortion even when accounting for the camera intrinsics.

Linked Images

Cycles Z pass to point cloud Eevee Z pass to point cloud

At this point I'd be assuming that Eevee is taking Z distance from the camera, whereas Cycles is calculating the length of the path before intersection, but I can't find any confirmation of these different behaviours.

$\endgroup$
2
  • $\begingroup$ Upon further searching, this post blender.stackexchange.com/questions/130970/… suggests that this is the result of the pinhole camera model used by the cycles renderer. As is pointed out; it's confusing that the two rendering methods have a different camera model; and this has persisted with the Introduction of Eevee $\endgroup$ – confused_perspective May 6 '20 at 11:17
  • $\begingroup$ This question may be duplicate of the previous, although the rendering engines being compared are different, the behaviour is the same. $\endgroup$ – confused_perspective May 6 '20 at 11:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.