"exsposure" is mainly to set the bright parts of the image and gamma is to set the dark parts. Try that and you will get much better images
It's a missunderstanding that gamma "should" be at 2.2
Also use Cone filtering at 1.5 + Fullscreen Grid AA 3 to get it fast and nice.
The image "the ring" in gallery is rendered with these settings.
Exposure and gamma and AA tip
-
Captain Obvious
- Posts: 165
- Joined: Mon Oct 10, 2005 6:15 pm
- Location: London, UK
-
Janusz Biela
- Posts: 3265
- Joined: Mon Mar 13, 2006 10:39 am
- Location: Finland
- Contact:
waiting for the tutorial also. 
by linear workflow, does that mean you have to inverse gamma correct all images/colors to gamma 0.4545 (in which the images will look extremely contrasty)? and later do a gamma 2.2 in kray to get it to look right. Is the purpose of this so that lights' inverse dist squared falloff will work correctly as it supposed to work?
by linear workflow, does that mean you have to inverse gamma correct all images/colors to gamma 0.4545 (in which the images will look extremely contrasty)? and later do a gamma 2.2 in kray to get it to look right. Is the purpose of this so that lights' inverse dist squared falloff will work correctly as it supposed to work?
Yes, exactly. When inv square lights are viewed at gamma 1.0 they will tend to falloff very quickly with a very bright (burned) hotspot. With gamma 2.2 they will look much more natural as can be seen here http://www.kraytracing.com/forum/viewto ... =2368#2368tajino wrote:waiting for the tutorial also.
by linear workflow, does that mean you have to inverse gamma correct all images/colors to gamma 0.4545 (in which the images will look extremely contrasty)? and later do a gamma 2.2 in kray to get it to look right. Is the purpose of this so that lights' inverse dist squared falloff will work correctly as it supposed to work?
Unfortunately gamma correcting every image and surface in a scene can be a very tedious process...
- Jure
it is the higher pixel filtering radius value that is good for less flickery animation, giving it a softer look and more pixel consistency thru frames.
Usually, scenes containing thin lines/dots will flicker thru frames with the default value of 0.5, which is a bit too small, the result might look razor sharp but it is bad for animation.
the reason for FSAA, it is more predictable, rather than gambling with adaptive edge detection. I would go for FSAA anytime, unless if the scene is made up of mostly flat, non-textured, non blur-reflection surfaces. With FSAA, i can lower the reflection rays as it will be compensated with each FSAA pass.
alternatively, i did try Box filter at 1.0, results are almost the same as cone, maybe only for the scene that i'm working on, got to do further testing on each filtering mode.
Talking about pixel-filtering mode, I've requested to G. some months ago about have an Off mode, no pixel filter at all. I know it might be similar with Box at 0.5, but it is the pixel filtering process that causes Kray to crash upon completing a render when ram is low. Which makes Kray incapable of doing large render (8000x6000) on a 2G ram limit (edit : without splitting up into limited region no-border segments).
Usually, scenes containing thin lines/dots will flicker thru frames with the default value of 0.5, which is a bit too small, the result might look razor sharp but it is bad for animation.
the reason for FSAA, it is more predictable, rather than gambling with adaptive edge detection. I would go for FSAA anytime, unless if the scene is made up of mostly flat, non-textured, non blur-reflection surfaces. With FSAA, i can lower the reflection rays as it will be compensated with each FSAA pass.
alternatively, i did try Box filter at 1.0, results are almost the same as cone, maybe only for the scene that i'm working on, got to do further testing on each filtering mode.
Talking about pixel-filtering mode, I've requested to G. some months ago about have an Off mode, no pixel filter at all. I know it might be similar with Box at 0.5, but it is the pixel filtering process that causes Kray to crash upon completing a render when ram is low. Which makes Kray incapable of doing large render (8000x6000) on a 2G ram limit (edit : without splitting up into limited region no-border segments).
What I do to streamline the process is to run a batch action script in photoshop of the image folder in my LW directory after I've built and surfaced my scene, so all images are gamma corrected for render time.jure wrote:Unfortunately gamma correcting every image and surface in a scene can be a very tedious process...
My PS action script is attached.
- Attachments
-
- Gamma_Correct.zip
- (660 Bytes) Downloaded 351 times
"Place your vote now for the Kraydar pixel filter"
Photoshop C2 &> can handle and retain 32bpc data.
The reason why I correct the original image opposed to modifying it in LW image edit is because....well.. I'm not that 100% sure, I think it's something to do with LW image editor having problems converting floating point data in the colour channel, but I'm not that sure, I've always done it this way and it works so I've not really needed another solution, maybe somebody could prove that I'm not going mad!
The reason why I correct the original image opposed to modifying it in LW image edit is because....well.. I'm not that 100% sure, I think it's something to do with LW image editor having problems converting floating point data in the colour channel, but I'm not that sure, I've always done it this way and it works so I've not really needed another solution, maybe somebody could prove that I'm not going mad!
"Place your vote now for the Kraydar pixel filter"