Displacement issue in V-ray
-
hi thomthom, thanks to that. i think i need to see now between the units. i remember also last time that you share how SU map the UV in inches, therefore i have resized my maps into 4X4 inches or 2x2 inches and it really works. as for the displace the graph was measured in pixel against cm graph so its a kinda based on the visual result and the pure black color has that ratio, the rest are not and these depends on the gradient from white to black. i myself still lokkin on the mapping issue and hope you could further shed some lights here to how displacemnent works. but the UV i beleive vray is using its own 1x1 inch unit.
-
@nomeradona said:
i myself still lokkin on the mapping issue and hope you could further shed some lights here to how displacemnent works.
What mapping issue?
-
@thomthom said:
@nomeradona said:
i myself still lokkin on the mapping issue and hope you could further shed some lights here to how displacemnent works.
What mapping issue?
how the displacement maps work in Vray. How the gray scale will be displace. its a kinda small test im doing here. i am setting 11 colors from white to black (9 grays in between from white to black) i have also list down the RGB numbers. I am planning to measure how the displacement behave.i just plan base on visual, but i know you have your test in mathematical forms. so if you have some thoughts to shed lights on this, i will appreciate.
-
Seems to be exponential. Wonder if it's possible to make it linear. Need to experiment. Maybe ask Damien.
I noted that turning off the filtering of the displacement map made it smoother. When it was on MipMap I got some weird artefacts.
-
I was wondering if there was any of the other settings that would make V-Ray treat the grayscale values linear instead of exponential. But I've yet to find it.
-
hmm.... could it be because I use 24bit images... Wonder if HDRI images will differ in behaviour...
-
I was correct in my suspicion.
When I converted the 8bit/color image to 32bit/color the medium 8bit gray (RGB 128,128,128) did not convert to 32bit medium gray (RGB 0.5, 0.5, 0.5) Instead it came out as RGB 0.216, 0.216, 0.216.
When I created a new grayscale image. This time in 32bit per channel, going from 0.0 to 1.0 with 0.1 increment I got a steady displacement.
-
But I wonder why I got the strange streaks in the displacement. I saw them in the 8bit version before I turned of filtering...
-
@thomthom said:
But I wonder why I got the strange streaks in the displacement. I saw them in the 8bit version before I turned of filtering...
i guess that one could not be perfected, that is why blurrying a bit the displacement is a key one. but even you do this streaks will still be there. waht about vray max? any idea. -
Don't have V-Ray for Max.
-
i made a similar test, and i found out a lot of things how vray behave with the grayscale. its smilar with yours thom thom. i will post later my images..
but these are the following facts.
- that the pure white with the multiplier 1 will be displace one inch perpendicular to the face.
- vray works in inches. not centimeters
- if you want to remove artifacts, use non filter and add blur in your displacement map. I found using Normal map seems better. i can prove it yet. but the amount of detail is excellent.
see this post http://www.sketchucation.com/forums/scf/viewtopic.php?f=81&t=18336
-
ah ok. anyone? let me find also in the Filipino forum.
-
FYI, the reason the 8-bit image has a curve is because the default V-Ray settings is to apply a 2.2 gamma correction.
-
@thomthom said:
FYI, the reason the 8-bit image has a curve is because the default V-Ray settings is to apply a 2.2 gamma correction.
The 2.2 correction stems from the General Switches I think. You can override the displacement gamma for that material with the inverse gamma - in the case of 2.2 you would have to enter 0.4545 in order to make the 8-bit displacement linear.
-
ThomThom
Brilliant piece of info
Possible use as disp for roof tiles, if you can calculate the increase in height of the plane?
Advertisement