h13 deep exr 2.0 to nuke8

   6322   4   0
User Avatar
Member
74 posts
Joined: Dec. 2009
Offline
rendered a dcm .exr from h13. However, the .exr is black in nuke8v1. In mplay, the .exr is not black.

In nuke8, the DeepSample node could display sample data.

Using iinfo dcm.exr, the WorldToCamera and WorldToNDC matrices are all zero. Is it a bug?

File ./dcm.exr (OpenEXR format)
Resolution 1920 x 1080

Pz
Data 32 bit float
Color Model Single Channel
Channel Depth 32 x 1 (32 bpp)

Of (r,g,b)
Data 16 bit float
Color Model RGB
Channel Depth 16 x 3 (48 bpp)

C (r,g,b)
Data 16 bit float
Color Model RGB
Channel Depth 16 x 3 (48 bpp)

A
Data 16 bit float
Color Model Single Channel
Channel Depth 16 x 1 (16 bpp)

Depth-Complexity
Data 16 bit integer
Color Model Single Channel
Channel Depth 16 x 1 (16 bpp)

Format Specific Information
Image Type Deep
Compression zips
Comment none
Created 20140129 173713
Depth Interpolation discrete
Samples Composited true
WorldToCamera [ 0, 0, 0, 0,
0, 0, 0, 0,
0, 0, 0, 0,
0, 0, 0, 0]
WorldToNDC [ 0, 0, 0, 0,
0, 0, 0, 0,
0, 0, 0, 0,
0, 0, 0, 0]
https://www.linkedin.com/in/kelvincai/ [www.linkedin.com]
User Avatar
Member
129 posts
Joined: Jan. 2013
Offline
HI,it works only nuke 7 ,not nuke 8. Maby it's a bug. I think it will be fixed in next ver.
User Avatar
Member
74 posts
Joined: Dec. 2009
Offline
I don't think nuke 7 support exr 2.0 deep data. And h13 mantra render exr 2.0 deep data in dcm
https://www.linkedin.com/in/kelvincai/ [www.linkedin.com]
User Avatar
Member
4189 posts
Joined: June 2012
Offline
Nuke8.0v2 works fine now with Mantra dcm Exr2

kelvincai
I don't think nuke 7 support exr 2.0 deep data. And h13 mantra render exr 2.0 deep data in dcm

Nuke 7 works well with Exr2 using DeepRead nodes
User Avatar
Member
92 posts
Joined: Aug. 2010
Offline
hey kelvin! yeah 8.0v1 deep is busted, so roll with something else.

Since you're getting into it, you may be interested in a couple tickets I filed related to Mantra's exr2:

SESI-16649 - store Zfront/Zback in deep exr images
Looking at volumetric deep samples via DeepSample in Nuke, there are values for “deep front”, but “deep back” is zero. While it might be ok to make the assumption that front=back for a hard-surface, you can't assume that deep samples are “tidy” (in-order and non-overlapping) for volumetric samples.

SESI-16644 - enable/disable toggle on deep raster planes
Add a toggle on deep raster planes to enable/disable it for DCM export; we would like to specify which channels get stored in the deep file.
  • Quick Links