Ahh right, so /tmp/rvLookLutTMP.csp is where I'm temporarily writing out the 3D only .csp before I read it back in and add the 1D shaper lut at the head, which then gets written out to your specified location.. It doesn't need to get included as its created mid way through the script (although you could point it to a different location if you wanted to some some reason).
The prelut uses the Nuke standard log2lin: (pow(10,(1023*x-685)/300)-.0108)/(1-.0108) But you can swap that out for whatever you want by changing the expression on the master channel in "1D Prelut" section of the tool. (or paste an existing curve in there). If you have an existing Lut you want to rebake just pop a Vectorfield or OCIOFileTransform node inside the group where the Grade and Saturation nodes are (after the switch, before the dot) -Alex On Wed, Jan 23, 2013 at 8:47 AM, Dan Rosen <[email protected]> wrote: > Hi Alex, > > I've got your tool working, though I have some questions. Is it possible > to set this up to feed it a log LUT as 3DL or other for the main lut part? > I don't see where I can do that in your current Group/Grizmo. It appears > that the tool is looking at "/tmp/rvLookLutTMP.csp" (lines 17 to the end), > but that file isn't included in your dropbox. I have pasted my own 3DL into > the main part of a test file and that works fine, but it'd be cool to get > the tool to do that. > > On a different, but somewhat related note, I always see a slight bump in > brightness in comparing a 3DL lut applied to a log image versus a CSP lut > applied to a linear image. I assume that the "shaper 1D" LUT is close, but > an approximation. I do have the correct equation for our Lin2Log, so maybe > some of the difference is coming from remapping the Log2Lin as a 1D lut. > Would that make the difference? > > thx, > Dan > > On Tue, Jan 15, 2013 at 7:02 PM, Alex - <[email protected]> wrote: > >> Hmm.. >> >> Try this version: >> https://www.dropbox.com/s/u3ykx7b8b92k50e/cinespaceLutGen_v002.nk >> >> Basically the same, but strips a few Linux specific font references. >> Working fine here on a stock instance of 6.3v8 under Linux. >> >> Will try it on a Mac at home later. >> >> Alex Fry >> >> >> On Wed, Jan 16, 2013 at 12:47 PM, Dan Rosen <[email protected]>wrote: >> >>> Hey Alex, >>> >>> Thanks. This looks promising. I'm getting some errors when I cut and >>> paste it into Nuke 6.3v8 on OSX 10.6.8. >>> >>> Disk cache /var/tmp/nuke-u1055/??: 482MB (1% of 51200MB) used in 59 >>> files. >>> [16:55.39] ERROR: ColorLookup1.lut: >>> [16:55.39] ERROR: ColorLookup1.lut: recursive expression in >>> (1/0.18)*(pow(10,(1023*x-444)/(1/(.002/0.45)))-0.0106332657)/(1-0.0106332657) >>> Traceback (most recent call last): >>> File >>> "/Applications/Nuke6.3v8/Nuke6.3v8.app/Contents/MacOS/plugins/nuke/callbacks.py", >>> line 161, in autolabel >>> s = f[0](*f[1],**f[2]) >>> File >>> "/Applications/Nuke6.3v8/Nuke6.3v8.app/Contents/MacOS/plugins/autolabel.py", >>> line 26, in autolabel >>> ind = >>> nuke.expression("(keys?1:0)+(has_expression?2:0)+(clones?8:0)+(viewsplit?32:0)") >>> ValueError: cannot convert to a floating point value >>> [17:39.35] ERROR: CineSpace3D_LutGenerator.lutRed: cannot convert to a >>> floating point value >>> [17:39.35] ERROR: CineSpace3D_LutGenerator.lutGreen: cannot convert to a >>> floating point value >>> [17:39.35] ERROR: CineSpace3D_LutGenerator.lutBlue: cannot convert to a >>> floating point value >>> [17:39.35] ERROR: CineSpace3D_LutGenerator.lutDestination: cannot >>> convert to a floating point value >>> Traceback (most recent call last): >>> File >>> "/Applications/Nuke6.3v8/Nuke6.3v8.app/Contents/MacOS/plugins/nuke/callbacks.py", >>> line 161, in autolabel >>> s = f[0](*f[1],**f[2]) >>> File >>> "/Applications/Nuke6.3v8/Nuke6.3v8.app/Contents/MacOS/plugins/autolabel.py", >>> line 26, in autolabel >>> ind = >>> nuke.expression("(keys?1:0)+(has_expression?2:0)+(clones?8:0)+(viewsplit?32:0)") >>> ValueError: cannot convert to a floating point value >>> [17:39.35] ERROR: VIEWER_INPUT.Switch_inputCspace1.which: cannot convert >>> to a floating point value >>> [17:39.35] ERROR: VIEWER_INPUT.Switch_aspectToggle.which: cannot convert >>> to a floating point value >>> [17:39.35] ERROR: VIEWER_INPUT.Aspect_1v4_GROUP.Switch1.which: cannot >>> convert to a floating point value >>> [17:39.35] ERROR: VIEWER_INPUT.Aspect_1v4_GROUP.Mask1.box: cannot >>> convert to a floating point value >>> [17:39.35] ERROR: VIEWER_INPUT.Aspect_1v4_GROUP.Mask1.reformat: cannot >>> convert to a floating point value >>> [17:39.35] ERROR: VIEWER_INPUT.Switch_bPointToggle.which: cannot convert >>> to a floating point value >>> [17:39.35] ERROR: VIEWER_INPUT.Switch_blurToggle.which: cannot convert >>> to a floating point value >>> [17:39.35] ERROR: VIEWER_INPUT.Switch_satToggle.which: cannot convert to >>> a floating point value >>> [17:39.35] ERROR: VIEWER_INPUT.FilterErode2.size: cannot convert to a >>> floating point value >>> [17:39.35] ERROR: VIEWER_INPUT.Add2.value: cannot convert to a floating >>> point value >>> [17:39.35] ERROR: VIEWER_INPUT.Switch_inputCspace.which: cannot convert >>> to a floating point value >>> Traceback (most recent call last): >>> File >>> "/Applications/Nuke6.3v8/Nuke6.3v8.app/Contents/MacOS/plugins/nuke/callbacks.py", >>> line 161, in autolabel >>> s = f[0](*f[1],**f[2]) >>> File >>> "/Applications/Nuke6.3v8/Nuke6.3v8.app/Contents/MacOS/plugins/autolabel.py", >>> line 26, in autolabel >>> ind = >>> nuke.expression("(keys?1:0)+(has_expression?2:0)+(clones?8:0)+(viewsplit?32:0)") >>> ValueError: cannot convert to a floating point value >>> [17:40.56] ERROR: CineSpace1D_LutGenerator.lutRed: cannot convert to a >>> floating point value >>> [17:40.56] ERROR: CineSpace1D_LutGenerator.lutGreen: cannot convert to a >>> floating point value >>> [17:40.56] ERROR: CineSpace1D_LutGenerator.lutBlue: cannot convert to a >>> floating point value >>> Traceback (most recent call last): >>> File >>> "/Applications/Nuke6.3v8/Nuke6.3v8.app/Contents/MacOS/plugins/nuke/callbacks.py", >>> line 161, in autolabel >>> s = f[0](*f[1],**f[2]) >>> File >>> "/Applications/Nuke6.3v8/Nuke6.3v8.app/Contents/MacOS/plugins/autolabel.py", >>> line 26, in autolabel >>> ind = >>> nuke.expression("(keys?1:0)+(has_expression?2:0)+(clones?8:0)+(viewsplit?32:0)") >>> ValueError: cannot convert to a floating point value >>> >>> >>> On Tue, Jan 15, 2013 at 4:00 PM, Alex - <[email protected]> wrote: >>> >>>> I ran into the same problem, so I made these: >>>> https://www.dropbox.com/s/ykzzlc6sm9hi7ji/cinespaceLutGen_v001.nk >>>> >>>> there are two nodes, one for making pure 1D .csp luts, another for >>>> making 3D luts with a 1D prelut. >>>> There are some basic grade controls on the face of the 3D lut >>>> generator, but you can open it up and put whatever random transforms you >>>> want in there. >>>> >>>> -Alex >>>> >>>> >>>> On Wed, Jan 16, 2013 at 7:25 AM, Dan Rosen <[email protected]>wrote: >>>> >>>>> Hello, >>>>> >>>>> I am trying to use GenerateLUT to author csp files for linear light >>>>> images to display conversion. We also have a custom LogLin so I'd like to >>>>> set the shaper lut to something custom with a file or values there. I'm >>>>> not >>>>> sure if the Nuke node requires that I have a config.ocio file or not, but >>>>> I >>>>> cannot seem to figure it out. I've looked in the docs and don't see it >>>>> there either. Any help is much appreciated. >>>>> >>>>> Thanks, >>>>> Dan >>>>> >>>>> _______________________________________________ >>>>> 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 >>>> >>> >>> >>> _______________________________________________ >>> 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 >> > > > _______________________________________________ > 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
