Nuke to Flint Workflow with EXR files

[Invader777]

Been wondering for a little while how to get this one up and running easily.


Flint (For some bizzare reason only know to Autodesk) exports EXR files in the SRGB colourspace. I can import these into Nuke and apply a lookup in the read node to correct the SRGB files back to Linear. When I import these (Linear) files back into Flint they appear to have darkened Gamma (As I expected.) But, If in my Nuke write node, I convert back to SRGB colourspace, then Flint refuses to open the file at all. (I could work in SRGB viewer colourspace in Nuke, but as it's a hatchetjob of the workflow - I'm not going to.).

Now, I was under the impression that EXR was by default, a Linear workspace (I know it CAN use other lookups - but Linear is superior)

Has anyone got this full EXR workflow up and running between FFI and Nuke?

I am eager to work in a half-float workspace, rather than continue to use (SRGB encoded) TIFF sequences.

(I used to use this workflow in shake - but R.I.P. and may apple choke on the iphone)

[beaker]

As usual their implementation is half assed Flint/Flame only supports none/rle compressed Exr files(this might have changed in the newest version). Why they can't at least support Zip compression is beyond me.

As for colorspace, yes, you have to burn in srgb into the file when going back and fourth. Not a big issue since Nuke supports it. Just make sure you use the "colorspace" node in Nuke with the colorspace in the Read node set to "linear"(<5.1) or "none"(>5.2x). This is because Nuke's read node clamps at 1.0 with srgb files and the colorspace node does not.

[rohitm]

try setting the flint viewer to LOG, then it works fine..am assuming u created a 3dLUT inside NUKE and are loading it into flint to view/work on the same sequence. true 3Dluts from lustre have no problems by the way.. as far as i know there are no standards defined for LUTS as of now ? but maybe im wrong.

Been wondering for a little while how to get this one up and running easily.


Flint (For some bizzare reason only know to Autodesk) exports EXR files in the SRGB colourspace. I can import these into Nuke and apply a lookup in the read node to correct the SRGB files back to Linear. When I import these (Linear) files back into Flint they appear to have darkened Gamma (As I expected.) But, If in my Nuke write node, I convert back to SRGB colourspace, then Flint refuses to open the file at all. (I could work in SRGB viewer colourspace in Nuke, but as it's a hatchetjob of the workflow - I'm not going to.).

Now, I was under the impression that EXR was by default, a Linear workspace (I know it CAN use other lookups - but Linear is superior)

Has anyone got this full EXR workflow up and running between FFI and Nuke?

I am eager to work in a half-float workspace, rather than continue to use (SRGB encoded) TIFF sequences.

(I used to use this workflow in shake - but R.I.P. and may apple choke on the iphone)

[Diogo Girondi]

Personally due to their OpenEXR implementation I would just stick with another file format. OpenEXR files make much more sense when the apps you're working do allow layered and linear encoded exr files. Not to mention the compression schemes limitations as Deke has pointed out.

But since I've been away from it for quite sometime perhaps all of those have changed in it's latest incarnations.

[rohitm]

nuke > flame i would simply work with .iff files.
even though flame can use nukes embedded EXR channels, it cant use them the same way as nuke does so whats the point of using EXRs from nuke in flame anyways. 32bit .iffs are good enough.

Personally due to their OpenEXR implementation I would just stick with another file format. OpenEXR files make much more sense when the apps you're working do allow layered and linear encoded exr files. Not to mention the compression schemes limitations as Deke has pointed out.

But since I've been away from it for quite sometime perhaps all of those have changed in it's latest incarnations.

[Invader777]

Truth be told - I'm quite happy with Beaker's EXR solution. I never use any compression anyway. I'm also quite happy with .Iff's. But it makes more sense to me to render to only one file format for my precomps and my final renders. EXR offers a little more flexibility than .IFF. - especially in rendering outside of the box - that is pretty cool. the 1024 layers thing is good for nuke - but I only end up rendering RGBA for flint most of the time.
Thanks Beaker.