search

cameramanben / LUTCalc

Web App for generating 1D and 3D Lookup Tables (LUTs) for video cameras that shoot log gammas.
http://www.lutcalc.net
GNU General Public License v2.0
249 stars 51 forks source link

cube file precision #14

Closed gtphil1 closed 1 year ago

gtphil1 commented 4 years ago

Hi, I'm running into precision limits when converting HLG to linear (Display HLG OOTF (Normalised)). Example file attached. It looks like values are limited to 6 decimal places in cube output. This LUT outputs really tiny values, so it ends up with banding and a lot of values rounded to zero. Would it be possible to use floating point exponential notation instead?

For context, I'm trying to work with HLG footage from the Fuji X-T3 in FCPX, and want to convert to linear space for color edits and other effects. I've tried applying this LUT and its reverse as a "no-op", and a lot of color definition is lost.

Thanks!

HLG_Linear.cube.txt

cameramanben commented 4 years ago

Hi gtphil1,

I'm just adding in a precision option in the settings - so you can set to arbitrary fixed-point decimal places,

Ben

cameramanben commented 3 years ago

Hi gtphil1,

sorry for my slowness - a few things meant that i had trouble pushing updates.

I've pushed some updates to a 'testing' version on the website: LUTCalc 3.3 Beta 1

this includes 'Decimal LUT Precision' under settings - the default is 8 decimal places, but you can go as far as 128 if you are so inclined ;-)

I'll spend some time checking I haven't done anything stupid with other additions and then should have it updated soon in the main online version and Mac app.

Thanks,

Ben