We recently switched to GPU Next at the office (coming from 3.2), and I have run into some issues.
1) Previously, you could render a z-depth pass using the "z-depth use camera clip boundaries" option in the Render Elements tab, and turn the "Clip Manually" function off in your camera so no geometry is clipped, but the values are passed to the z-depth element. However, now with GPU Next, if you turn the Clip Manually option off, it renders black (because clipping is off?). This is a real downer because it's super easy to animate the clipping planes to have the depth channel change focus.
Because of the above issue, I'm rendering my z-depth & velocity elements in a second pass using CPU Next. However, now I have run into an issue with that.
2) I am having problems getting a .EXR file with a z-depth element that has no alpha channel applied. In After Effects, I interpret my footage as alpha ignored and extract the Z channel, but it still appears to have an alpha applied (see attached picture, if I flip the transparency grid on, there's a phantom transparent background, but objects are not visible through it {pink solid in background} ). GPU Next creates a z-depth element correctly, but because of issue #1, I can't really use that.
Any thoughts? Bugs? User error?
Thanks!
1) Previously, you could render a z-depth pass using the "z-depth use camera clip boundaries" option in the Render Elements tab, and turn the "Clip Manually" function off in your camera so no geometry is clipped, but the values are passed to the z-depth element. However, now with GPU Next, if you turn the Clip Manually option off, it renders black (because clipping is off?). This is a real downer because it's super easy to animate the clipping planes to have the depth channel change focus.
Because of the above issue, I'm rendering my z-depth & velocity elements in a second pass using CPU Next. However, now I have run into an issue with that.
2) I am having problems getting a .EXR file with a z-depth element that has no alpha channel applied. In After Effects, I interpret my footage as alpha ignored and extract the Z channel, but it still appears to have an alpha applied (see attached picture, if I flip the transparency grid on, there's a phantom transparent background, but objects are not visible through it {pink solid in background} ). GPU Next creates a z-depth element correctly, but because of issue #1, I can't really use that.
Any thoughts? Bugs? User error?
Thanks!
Comment