Correct point light attributes #211
Conversation
|
sorry for being a little late here. See this for how light attenuation works in Ogre: however I cannot find anything on how light attenuation works in Blender mathematically.. |
|
In the end I found a formula in the Wiki and used that. Then I had to adjust 10x for some reason for it to look more closely to Blender. However Blenders shaders are of course different than OGREs |
| light_range = ob.data.cutoff_distance * 10 | ||
| if light_range == 0: | ||
| light_range = 0.001 | ||
| a.setAttribute('range', light_range) |
There was a problem hiding this comment.
this should be the same as cutoff_distance
| a.setAttribute('range', light_range) | ||
| a.setAttribute('constant', '1.0') | ||
| a.setAttribute('linear', '%6f' % (4.5 / light_range)) | ||
| a.setAttribute('quadratic', '%6f' % (75.0 / (light_range * light_range))) |
There was a problem hiding this comment.
this should be 2/distance*distance if I read the blender docs and this correctly: http://learnwebgl.brown37.net/09_lights/lights_attenuation.html
the linear parameter should be 0
There was a problem hiding this comment.
Thanks, I'll look into it.
There was a problem hiding this comment.
There is very little difference, but I think that its better using 1/d^2.
-
With the formula: 1+4,5 / (10 *d) + 7,5 / (10 * d)^2
-
Using the formula proposed in the site (1+ 1 / d^2):
-
Blender:
There was a problem hiding this comment.
there is a much clearer test that you can use:
There was a problem hiding this comment.
part of the issue is that powerScale in ogre is not the same as ob.data.energy:
forcing power scale to 1.0, instead of 1000 (energy) gives you:
the remaining issue is the ambient light setting, that ogre-meshviewer currently ignores (which is ok I think) and I guess blender does something special at the cutoff distance. some sort of fade-out whereas Ogre just does a hard cut.
There was a problem hiding this comment.
There seems to be some sort of ambient lighting in Blender.
In OGRE I got a hard cut at 5 "meters" which was the range or distance.
In my case I was using shaders from https://learnopengl.com/
But yes, your idea of using a plane is much better.
Based on the information here: https://wiki.ogre3d.org/-Point+Light+Attenuation
I changed the calculation for the Point Light parameters.
Had to introduce 10x factor because otherwise it does not look like in Blender.
This also solves the crash reported in #209