I think if you have a .cube with output values over 1.0, you should see those values in AE. Make sure your project is in 32-bit mode.
If you want to do a comparison, you can try the OpenColorIO plug-in, which can also read in a .cube file.
Thank you for your help. When I try that I get an error message "The specified transform file could not be loaded. Malformed color triples specified in Iridas.cube lut:'LUT_1D_INPUT_RANGE -0.01(ect) 54.8(ect)'." (ect being the rest of the numbers past the decimal point.) I've come across this problem before in dealing with HDR productions and and it seems like it become more of an issue as time goes along.
Is that message from AE or OpenColorIO? I guess your .cube has some sort of LUT_1D_INPUT_RANGE property to adjust the input. If you open the file in a text editor, what does that line say?
Sounds like it's something like this:
LUT_1D_INPUT_RANGE -0.01234 54.890123
But maybe it should be RGB triples like this:
LUT_1D_INPUT_RANGE -0.01234 -0.01234 -0.01234 54.890123 54.890123 54.890123
What program made the LUT?
Maybe the easiest thing would be to delete that line from your LUT and put those values in the Input Black/White fields of a Levels before the LUT and then again in Output Black/White after the LUT.
I think the message was technically through AE but when I was using OpenColorIO. Modifying the LUT to remove the input range property allows me to import it, however the suggestion of putting the range values into the input B/W fields of a levels adjustment before the LUT and then again in the output values of a levels adjustment after the LUT does not produce the correct result.
I'm comparing a still with the LUT applied in Davinci Resolve to the LUT inside AE. I did check the source plates in both programs and they match, as well as a baked reference QT against the raw plate with provided LUTs applied through Resolve.
Tripling the values causes the same crashes/rejections that the single values caused.
I initially thought it was working once I was able to load the LUT, but realized later once I compared the images. I believed I may have incorrectly marked your previous suggestion as correct.
Sorry about that.