Date
1 - 4 of 4
Building config.ocio from provided LUTs
robin...@...
Hi, I'm new to OCIO and this forum, and I have what I think is probably a pretty common issue, although after searching for a solution for a couple days, I haven't come up with anything. My problem is that I am receiving pre-generated LUTs from vendors, mostly in the form of .cube and .spi1d files, and I would like to integrate them into our custom config.ocio. Is there an established way to extract the required information from these files so I can populate the colorspace definitions in my config? Most of the fields can be determined, but bitdepth and interpolation are often not defined. Is there a way to figure this out from the information in the LUT? |
|
Michel Lerenard <micheli...@...>
Hi
toggle quoted message
Show quoted text
as far as I know, no. 'allocation' too cannot be deduced. The only solution i can think of is having a file attached to the LUT specifying these parameters. On 11/07/2013 02:55 AM, robin...@... wrote:
Hi, |
|
dbr/Ben <dbr....@...>
Neither of those can really be determined automatically from the LUT file alone Determining the interpolation requires visually inspecting the results of the LUT on various images.. but using linear as a default is probably a good approach if you have to You could potentially determine the bitdepth of the LUT by checking the range of the values (for integer-based LUT formats like 3dl), but not in a general sense - e.g a LUT which stores values as floating point (e.g spi1d I think, or definitely the Cinespace CSP format) On 07/11/2013, at 12:25 PM, robin...@... wrote:
|
|
Robin Dutta <robin...@...>
I feared as much, but I'm surprised that with all the flexibility and
uniformity that OCIO provides, there's no method to share LUTs. We've asked the vendors to provide their config.ocios along with their LUTs, but they've been unwilling to do so. All I can figure out so far is the Allocationvars from an spi1d. I was also hoping that I might be able to figure out the bit depth based off the max number of decimal places their LUT values have, but am not sure if this is a sound approach. My only other recourse is to have someone make a best guess on the fields which I'm unsure of. On Friday, November 8, 2013 5:12:02 AM UTC-8, dbr/Ben wrote:
|
|