Jump to content

  • Log In with Google      Sign In   
  • Create Account


#ActualMJP

Posted 14 January 2013 - 05:34 PM

One thing some games do is they will "normalize" their reflection probes by dividing all texels by the intensity of the brightest texel. Then at run time they will multiply the reflection map value by the intensity of the baked diffuse lighting (usually either light maps or SH environment probes). This helps the probes "fit in" better with varying lighting conditions, especially if your probes are sparse.

You can also potentially make your reflections a lot better in all conditions by pre-baking some AO. This can be especially useful if you're using a fresnel term, since it tends to produce a rim-lighting-esque effect on meshes if you don't apply any visibility to the sampled reflections.


#2MJP

Posted 14 January 2013 - 05:32 PM

One thing some games do is they will "normalize" their reflection probes by dividing all texels by the intensity of the brightest texel. Then at run time they will multiply the reflection map value by the intensity of the baked diffuse lighting (usually either light maps or SH environment probes). This helps the probes "fit in" better with varying lighting conditions, especially if your probes are sparse.


#1MJP

Posted 14 January 2013 - 05:32 PM

One thing some games do is they will "normalize" their reflection probes by dividing all texels by the intensity of the brightest texel. Then at run time they will multiply the reflection map value by the intensity of the baked diffuse lighting (usually either light maps or SH environment probes). This helps the probes "fit in" better with varying lighting conditions, especially if your probes are sparse.


PARTNERS