Some thoughts/guesses: >From the command line you posted, it seems like the lut you have is trying to go from linear to your viewing space, correct?
Is it possible your issues are stemming from a) linear luts coming from linear space generally only get applied to the data between 0 and 1 (because, where do you stop?) b) it takes a lot of values to encode linear space linearly, and you may be seeing significant quantization. The difference between the lut in xsi and in nuke could possibly be due to different interpolation modes between the two programs. Maybe Nuke is doing a better job approximating a curved response between data points in the lut? I'm just speculating. I'm not sure why the lut would look overexposed in xsi. Maybe it's the same quantization thing as b? Can you confirm what colorspaces the original 3dl is mapping between? On Mon, Mar 24, 2014 at 12:16 PM, Simon van de Lagemaat < [email protected]> wrote: > Here's what my comp guy is saying. Seems like most people who have success > aren't converting their luts but instead are creating them from scratch. > > "We're converting a 3dl lut with ociobakelut > > We're using ociobakelut with config-free baking. > > We want to convert it to a .csp file to use in XSI > > current command is > > ociobakelut --lut lut.3dl --format cinespace lut.csp > > However when we load the lut.csp into soft without pre-conditioning we > don't get a match? Colour kinda looks ballpark but it looks over exposed > > Loading the .csp back into nuke and it matches with the .3dl ?" > > > On Mon, Mar 24, 2014 at 10:52 AM, Andy Jones <[email protected]> wrote: > >> Maybe you need to add a shaperspace? >> >> ociobakelut.exe --iconfig <ocio_config_path> --format cinespace >> --inputspace <inputspace> shaperspace lg10 --outputspace <outputspace> >> --looks <looks> <output_path> >> >> >> On Mon, Mar 24, 2014 at 8:58 AM, Simon van de Lagemaat < >> [email protected]> wrote: >> >>> 2013 w the sp >>> >>> >>> On Monday, March 24, 2014, Jens Lindgren <[email protected]> >>> wrote: >>> >>>> What version of Softimage are you using? >>>> >>>> /Jens >>>> >>>> >>>> On Sat, Mar 22, 2014 at 12:11 AM, Simon van de Lagemaat < >>>> [email protected]> wrote: >>>> >>>>> Trying to get a show lut going in the region render and we're getting >>>>> consistent clipping in the highlights no matter what type of lut (csp and >>>>> cube) we try and no amount of preconditioning is helping... is this just a >>>>> case of the region render not being fp? Is there a possible fix for this >>>>> I'm not aware of? >>>>> >>>> >>>> >>>> >>>> -- >>>> Jens Lindgren >>>> -------------------------- >>>> Lead Technical Director >>>> Magoo 3D Studios <http://www.magoo3dstudios.com/> >>>> >>> >> >
