« first day (235 days earlier)      last day (733 days later) » 

cegaton
cegaton
03:23
sRGB EOTF 0.18 mapped to 0.4614 White Clip roughly 2.5 stops over medium gray.
user image
Log No View.
user image
Log Low Contrast. S shaped curve.
user image
Log Basic Contrast
user image
Log High Contrast
user image
 
4 hours later…
troy_s
troy_s
07:41
@cegaton Very cool. They are all S transforms, just varying shapes
 
9 hours later…
troy_s
troy_s
16:26
@cegaton How the hell are you generating those curved nonlinear results???
tynaud
tynaud
16:48
From top to bottom, using basic contrast : false color , carbon fiber coated, plastic, silver, gold, copper, aluminum. I should make aluminum and silver more different, and should try to get more expo with the rim light, as 2 or 3 stops...
user image
I am not sure about middle gray. Diffuse with some 0.18 (rgb) in the gray (expo 0) should be at 0.5 or 0.6 ?
cegaton
cegaton
17:42
@troy_s simple patches, one stop difference between them (except the brightest, that is only 1/2 over to reach 16.291)
 
4 hours later…
troy_s
troy_s
21:54
@tynaud Middle grey in the new Filmic set maps 0.18 to precisely 0.5
@tynaud The Log Encode maps 0.18 to 0.6, as it did previously.
(2^6.5) * 0.18
Gah
Should be something like 10/16.5
Idiot.
@cegaton But how are you getting the image log-encoded? Those scopes are showing log. That is just on the view I guess.
Greets @D.Va
Banished from chat by moderator.
cegaton
cegaton
22:26
@troy_s Nothing special. Just changing the views. The scopes follow...
troy_s
troy_s
Hrm.
Post view and pre view transforms. I rarely use those scopes, but similar concepts there.
cegaton
cegaton
@troy_s I find them useful in this case. It shows what the different contrast views are working.
troy_s
troy_s
Yes. Absolutely in terms of the display referred land.
As the data is pretty much all or nothing at that point.
A 256 or 1024 step gradient would be excellent.
Great way to show the shape of the curves!
Not sure how to generate that in Blender.
cegaton
cegaton
@troy_s indeed. At some point I tried to generate a simple ramp going form 0 to 16.291, but I can't get it correctly.
troy_s
troy_s
For the scopes test, a simple 256 or 1024 would suffice, as it is post view transform.
@cegaton Think you can help me out here?
He's looking for examples, and I'm slow as piss with them. developer.blender.org/T47212
I'd love to get the damn thing actually be a "Luminance" matte, not "Some random whacky primaries from some misinformed developer Matte"

« first day (235 days earlier)      last day (733 days later) »