Hi Håkan, thanks for following up on this. You're absolutely correct, in my initial post I hadn't set the illuminant to D65. Also, I realize the ACES->Alexa and ACES->sRGB was a bit confusing. The whole reason I became suspicious of the colorspace node was that I had seen a ACES->sRGB matrix that didn't match, and then I became worried that the the ACES->Alexa was off as well (I still haven't found this matrix by the way).
I didn't follow up on the thread but I actually used Ben's script to do the ACES->Alexa calculation, and that matched Nuke's colorspace (fairly close), as well as the official paper (LogC - Usage in VFX <http://www.arri.com/camera/alexa/downloads/>). Good to see your calculation as well! It should be pretty safe to assume that the colorspace works correctly :). Best regards, Simon ------------------------------- Simon Björk Compositor/TD +46 (0)70-2859503 www.bjorkvisuals.com 2014-10-29 10:07 GMT+01:00 Håkan Blomdahl <[email protected]>: > Also, sorry if I am confused: you write ACES->Alexa and then mention > ACES->sRGB... > > Just to test Nuke ACES->Alexa matrix with some external calculations, I > have some experimental python code here based on bruce lindbloom formulas > and using NumPy: > > rgbw_ACES = colormatrix.RGBW() > rgbw_ACES.set_rgbw_xy( 0.73470, 0.26530, 0.00000, 1.00000, 0.00010, > -0.07700, 0.32168, 0.33767 ) > rgbw_Alexa = colormatrix.RGBW() > rgbw_Alexa.set_rgbw_xy( 0.6840, 0.3130, 0.2210, 0.8480, 0.0861, -0.1020, > 0.3127, 0.3290 ) > > # Generate Matrixes to and from XYZ > M_ACES_to_XYZ = rgbw_ACES.matrix_to_XYZ() > M_Alexa_from_XYZ = rgbw_Alexa.matrix_from_XYZ() > > # Combine Matrixes to one: > M_ACES_to_Alexa = M_ACES_to_XYZ * M_Alexa_from_XYZ > print M_ACES_to_Alexa > > > And the result gets close enough: > My code: > 1.53820322 -0.3513648 -0.16686751 1.01681111 -0.12917464 0.09536237 > 0.05995383 -0.01124037 0.88068171 > Nuke: > 1.53818 -0.351359 -0.166865 1.01683 -0.129178 0.0953645 > 0.0599613 -0.0112418 0.880791 > > So the Nuke node seems fine. I have not got total control of floating > point precision, data types etc in my calculations, but it seems to be in > the ball park. > /Håkan > > > > On 29 October 2014 09:41, Håkan Blomdahl <[email protected]> wrote: > >> Hi! >> >> It looks like on the colorspace node you did, you only changed the >> primaries to ACES but left the Illuminant to D65 ( instead of ACES ). >> >> In which case the matrix will be: >> 1.53818 -0.351359 -0.166865 -0.129178 1.01683 0.0953645 -0.0112418 >> 0.0599613 0.880791 >> >> Could that be it? >> >> /Håkan >> >> >> >> On 20 October 2014 13:05, Simon Björk <[email protected]> wrote: >> >>> Deke: I was sure I that it was in this (very interesting) thread >>> <https://groups.google.com/forum/#!topic/academyaces/9b4VuqPcOHQ> I >>> read it, but I'm actually unable to find it at the moment. It was not an >>> official matrix. >>> >>> Ben: Thanks a lot for those links. I will see if I can get your script >>> working. Interesting about the color temperature by the way, I didn't know >>> that. >>> >>> Best regards, >>> Simon >>> >>> >>> >>> >>> >>> ------------------------------- >>> Simon Björk >>> Compositor/TD >>> >>> +46 (0)70-2859503 >>> www.bjorkvisuals.com >>> >>> 2014-10-20 4:44 GMT+02:00 Ben Dickson <[email protected]>: >>> >>>> > Can you point to the ACES to sRGB matrix you found? >>>> >>>> It'll just be the ACES -> XYZ multiplied with an XYZ->sRGB matrix. >>>> >>>> You could use a ColorMatrix node to convert from ACES->XYZ, then a >>>> second ColorMatrix to go from XYZ->sRGB using the appropriate matrix >>>> from, say, >>>> >>>> >>>> http://www.brucelindbloom.com/index.html?Eqn_RGB_XYZ_Matrix.html#WSMatrices >>>> >>>> >>>> > I've been playing with the colorspace node in Nuke to convert an >>>> > ACES file into AlexaV3LogC (primaries only). Anyone know if the >>>> > matrix used by the colorspace node is correct? >>>> >>>> Not sure. It might depend how accurate you want to be. If you look here: >>>> >>>> >>>> https://github.com/ampas/aces-dev/tree/master/transforms/ctl/idt/vendorSupplied/arri/alexa >>>> ..specifically this file: >>>> >>>> https://github.com/ampas/aces-dev/blob/master/transforms/ctl/idt/vendorSupplied/arri/alexa/AlexaParameters-2013-Nov-13/Alexa_aces_matrix.txt >>>> >>>> ..has different matrices for each colour temperature (for when >>>> converting from Alexa footage to ACES) >>>> >>>> More simply, if you just want to do basic conversion from ACES to sRGB >>>> gamut, then you could use this code: >>>> https://gist.github.com/dbr/24cfd1033c2d59f263e3 >>>> (as described above, by calculating the ACES->XYZ matrix then XYZ->your >>>> target gamut) >>>> >>>> >>>> On 20/10/14 12:01, Deke Kincaid wrote: >>>> > Can you point to the ACES to sRGB matrix you found? The matrices that >>>> > are published from the Academy are going from XYZ to ACES and vs >>>> versa. >>>> > >>>> > -- >>>> > Deke Kincaid >>>> > Creative Specialist >>>> > The Foundry >>>> > Skype: dekekincaid >>>> > Tel: (310) 399 4555 - Mobile: (310) 883 4313 >>>> > Web: www.thefoundry.co.uk <http://www.thefoundry.co.uk/> >>>> > Email: [email protected] <mailto:[email protected]> >>>> > >>>> > On Sun, Oct 19, 2014 at 5:18 AM, Simon Björk <[email protected] >>>> > <mailto:[email protected]>> wrote: >>>> > >>>> > I've been playing with the colorspace node in Nuke to convert an >>>> > ACES file into AlexaV3LogC (primaries only). Anyone know if the >>>> > matrix used by the colorspace node is correct? >>>> > >>>> > Colorspace matrix (ACES->AlexaV3LogC): >>>> > 1.53461874, -0.354506, -0.180112 >>>> > -0.128879, 1.02594, 0.102936 >>>> > -0.0112158, 0.0604984, 0.950717 >>>> > >>>> > The reason that I'm unsure is that I've seen a ACES->sRGB matrix >>>> on >>>> > another mailing list that does not match the values I find in >>>> Nuke. >>>> > >>>> > Best regards, >>>> > Simon >>>> > >>>> > ------------------------------- >>>> > Simon Björk >>>> > Compositor/TD >>>> > >>>> > +46 (0)70-2859503 <tel:%2B46%20%280%2970-2859503> >>>> > www.bjorkvisuals.com <http://www.bjorkvisuals.com> >>>> > >>>> > _______________________________________________ >>>> > Nuke-users mailing list >>>> > [email protected] >>>> > <mailto:[email protected]>, >>>> > http://forums.thefoundry.co.uk/ >>>> > >>>> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >>>> > >>>> > >>>> > >>>> > >>>> > _______________________________________________ >>>> > Nuke-users mailing list >>>> > [email protected], http://forums.thefoundry.co.uk/ >>>> > http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >>>> > >>>> >>>> -- >>>> ben dickson >>>> 2D TD | [email protected] >>>> rising sun pictures | www.rsp.com.au >>>> _______________________________________________ >>>> Nuke-users mailing list >>>> [email protected], http://forums.thefoundry.co.uk/ >>>> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >>>> >>> >>> >>> _______________________________________________ >>> Nuke-users mailing list >>> [email protected], http://forums.thefoundry.co.uk/ >>> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >>> >> >> >> >> -- >> Håkan Blomdahl >> Filmgate AB >> >> Järntorget 3 >> 413 04 Göteborg >> Sweden >> >> www.filmgate.se >> [email protected] >> +46 (0)31- 701 02 00 >> > > > > -- > Håkan Blomdahl > Filmgate AB > > Järntorget 3 > 413 04 Göteborg > Sweden > > www.filmgate.se > [email protected] > +46 (0)31- 701 02 00 > > _______________________________________________ > Nuke-users mailing list > [email protected], http://forums.thefoundry.co.uk/ > http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users >
_______________________________________________ Nuke-users mailing list [email protected], http://forums.thefoundry.co.uk/ http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
