Thanks Kevin thats very helpful.
Yes in display space you end up dealing at least with sRGB or DCI. When going from LOG to LIN i have no idea what the primaries are - or even if these are known when using negs…? I have a similar bunch of setups, similar to yours you have sent, from Alex Fry. The only issue I can see is that in this technique the final colorspace node (taking you to CIE) sets the primaries ( to whatever is in the list) and so you are effectivley defining it. Although nonetheless a very cool way to visualise LUTs. Here's my set up.. http://pastebin.com/V71Wpu37 <http://pastebin.com/V71Wpu37> N Neil Rögnvaldr Scholes Director | VFX Supervisor | Artist www.neilscholes.com <http://www.neilscholes.com/> +44 (0) 7977 456 197 > On 10 May 2016, at 10:18, Kevin Wheatley <kevin.wheat...@framestore.com> > wrote: > > On 7 May 2016 at 20:28, Neil Scholes <n...@uvfilms.co.uk> wrote: >> I have been using the OCIO SPI-VFX route, where you view through a film sRGB >> display LUT. So in a sense the primaries are viewed as sRGB... Or DCI is >> the other option. And I can jump between DCI and sRGB on my monitor.. Eizo >> CG245w. > > when you view an image through a 3D LUT tagged with sRGB/709 or with > 'P3', usually this is relating to the state of the output device you > are using, it says nothing about the incoming data. > > For instance, if you use the ARRI Lut generator to build you a LUT you > use the colour space setting within the *conversion* tab as in > "convert to", in all cases they assume Alexa wide gamut primaries as > input and adjust the output to work with a display calibrated in the > selected colour setting. > > In the case of the SPI luts I'm not sure exactly what they assume, but > the curving to me looks like a change of primaries somewhere in there, > but you can visualise what they do using something like the below and > viewing the results (try disable the gamma and colorspace nodes to > visualise in 3D), with them turned on and viewing the rtside will show > you something like a CIE 1931 plot (though the axis need flipping > about) Obviously if you want exactly what the ocio config does you > should feed that in. > > Kevin > > set cut_paste_input [stack 0] > version 9.0 v8 > CMSTestPattern { > inputs 0 > cube_size 33 > name CMSTestPattern2 > selected true > xpos -2324 > ypos -126 > } > set Nf18286f0 [stack 0] > OCIOFileTransform { > file spi_ocio_srgb_test.spi3d > working_space linear > name OCIOFileTransform1 > selected true > xpos -2324 > ypos -54 > } > Gamma { > value 0.4545454545 > name Gamma2 > selected true > xpos -2324 > ypos -28 > } > Colorspace { > colorspace_out CIE-Yxy > name Colorspace1 > selected true > xpos -2324 > ypos 10 > } > PositionToPoints2 { > display textured > render_mode textured > P_channel rgb > name PositionToPoints3 > selected true > xpos -2324 > ypos 36 > } > push $Nf18286f0 > OCIOFileTransform { > file colorworks_filmlg_to_p3.3dl > working_space linear > name OCIOFileTransform2 > selected true > xpos -2208 > ypos -53 > } > Gamma { > value 0.3846153846 > name Gamma1 > selected true > xpos -2208 > ypos -27 > } > Colorspace { > illuminant_in DCI-P3 > primary_in DCI-P3 > colorspace_out CIE-Yxy > name Colorspace2 > selected true > xpos -2208 > ypos 11 > } > PositionToPoints2 { > display textured > render_mode textured > P_channel rgb > name PositionToPoints4 > selected true > xpos -2208 > ypos 37 > } > > -- > *Framestore* | Kevin Wheatley | Head of Imaging > *London* - New York - Los Angeles - Montreal > > 19-23 Wells Street, London, W1T 3PQ > T: +44 (0) 20 7344 8000 > > framestore.com <http://www.framestore.com/> > Facebook <https://www.facebook.com/Framestore> > Twitter <http://twitter.com/Framestore> > _______________________________________________ > Nuke-users mailing list > Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/ > http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >
_______________________________________________ Nuke-users mailing list Nuke-users@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/ http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
