So Vectorfield would still be wrong? Or something in our .csp file? I'm
not sure I care about the nodes (I would care about our lut, however),
I'd just rather know what to trust. Sounds like OCIO is the way to go?
My example was a just a source image and looking at the 2 nodes side by
side, viewer gained up by 8, no viewer lut with linear image. Using 7.0v8
jrab
On 11/22/2013 01:26 AM, Ben Dickson wrote:
Hm, the known-difference between Vectorfield and OCIOFileTransform
appears to be fixed
There was a bug where the Vectorfield was definitely wrong - creating a
no-op CSP LUT with CMSTestPattern->GenerateLUT, then applying it with
the Vectorfield caused a luminance shift (when it should do nothing)
..but it's fixed as of 6.3v8
The slight difference looks more like interpolation difference
On 22/11/13 03:14, Diogo Girondi wrote:
There is a known gain difference between OCIO and Vectorfiled while
working with .csp LUTs. Supposedly OCIO is correct from what I recall.
This info is somewhere on the OCIO website.
Here we are trying to keep LUTs in either 3DL or Cube.
Cheers,
Diogo
On Thu, Nov 21, 2013 at 6:42 AM, John RA Benson<[email protected]
<mailto:[email protected]>> wrote:
I'm getting a slight difference in the bottom end with
OCIOFileTransform and Vectorfield:
<vect.jpg>
<ocio.jpg>
<diff.jpg>
I will admit, the viewer is boosted quite a bit to see this, and it
is super slight.
I doubt it really matters, but it's good to be aware there is a
difference if anyone gets tweaked. I'm assuming that using the
OCIOFileTransform would be the preferred method going forward?
cheers
JRAB
On 11/21/2013 12:49 AM, John Vanderbeck wrote:
Viewer is Half-Float.
Changing to OCIFileTransform does fix the issue, so I will go
forward with that. Thanks! Not too well versed on OCIO so I
didn't think to try that.
*John Vanderbeck*
*2D Pipeline TD*
*T:* +1 604 733 7030
1205 Melville Street, Vancouver, B.C, V6E 0A6, Canada.
primefocusworld.com<http://www.primefocusworld.com>
On Wed, Nov 20, 2013 at 3:12 PM, Ben Dickson
<[email protected]<mailto:[email protected]>> wrote:
If you're using a vaguely-recent version of Nuke, you could
use the
OCIOFileTransform node instead of the Vectorfield
On 21/11/13 09:05, John Vanderbeck wrote:
> Hey all,
>
> Anyone else run into this?
>
> In a standard Nuke script you can bring in a Vectorfield and
load a .csp
> LUT file and it works fine.
>
> Do the same thing as an automated ViewerProcess though and
the color is
> all wrong. Use a .cube file instead and it works.
>
> Problem is, while using .cube works, my color guys tell me
that the
> .cube file can't handle the proper transforms for what we
need on this show.
>
>
>
>
> *John Vanderbeck*
> *2D Pipeline TD*
> *T:* +1 604 733 7030<tel:%2B1%20604%20733%207030>
>
> 1205 Melville Street, Vancouver, B.C, V6E 0A6, Canada.
> primefocusworld.com<http://primefocusworld.com>
<http://www.primefocusworld.com>
>
>
>
> _______________________________________________
> Nuke-python mailing list
> [email protected]
<mailto:[email protected]>,
http://forums.thefoundry.co.uk/
>
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-python
>
--
ben dickson
2D TD | [email protected]<mailto:[email protected]>
rising sun pictures | www.rsp.com.au<http://www.rsp.com.au>
_______________________________________________
Nuke-python mailing list
[email protected]
<mailto:[email protected]>,
http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-python
_______________________________________________
Nuke-python mailing list
[email protected], http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-python
_______________________________________________
Nuke-python mailing list
[email protected], http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-python
_______________________________________________
Nuke-python mailing list
[email protected], http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-python
