dvlenk6 opened this issue on Oct 22, 2006 · 3 posts
dvlenk6 posted Sun, 22 October 2006 at 9:32 PM
These are purposefully simplistic models. The quality is not an issue, neither is the content of the model.
The first picture is IBL (with nothing but a LDRI, black with several grey scaled blotches around the edges).
The second is the same, with True Ambience enacted. All surfaces lack either specularity or reflectivity. If this is not thermal radiance, what is it? No other optical phenomenon can account for this.
Is it really so hard to believe that a raytracer developed in the same laboratory that developed the concepts of both raytracing and radiosity in the first place might have an algorithm in it that can closely approximate radiosity?
Let's be serious.
I feel that Bryce has never had the capability of recieving high quality global illumination before. It does now, so that at least part of the True Ambience alogrithm can be enacted.
Currently, I believe that Bryce's Diffuse inter-reflectivity simulation is hampered by:
A) the inability of the algorithm to return a [null], or used defined percentage, from areas that are recieving neither direct nor indirect illumination. Shadows fade out, dark areas appear to glow, just as before IBL was introduced.
b) A more accurate sampling method, i.e. particulate photon sampling, than standard distributed raytracing.
However, There are benefits to this approach over the enactment of Radiosity that I see.
Radiosity is fundamentally based entirely in global illuminatuion, it cannot simulate a positional optical phenomenon.
IBL + TA on the other hand has exhibited the ability to simulate both specular and reflective transmissions as well.
It will be interesting to see where this leads.
I am not really concerned with how any application other than Bryce handles this, nor any estimations about 'true' vs. 'fake', as all rendering processes are approximations, those opinions are meaningless. All that matters is how accurate the approximation can be made to be.
Friends don't let friends use booleans.
dvlenk6 posted Sun, 22 October 2006 at 9:41 PM
I'll try to post the scenefile again.
Friends don't let friends use booleans.
fpfrdn3 posted Sun, 22 October 2006 at 9:47 PM
I would be intrested to see the trade off in render times for IBL and/or TA that people will have vs. the quality of scene or lighting. Such as AgentSmith has done in some pics here. Also how everyone will try to implement it in Bryce 6,( which I don't have yet), because is does look good though. Yes, it will be interesting. 