Bryce5.com

Registered users

» Forgot password
» Registration

Random image

Bombardment
Comments: 4
connorzelinsky

16.04.2024, 08:47

Home Page

Google ads below

Global Illumination Study Random Objects

Global Illumination Study Random Objects
Description:	I figured that with hundreds of lights on the walls and celing and floors that I should be able to produce very high levels of realism. The objects in the room are being lit from all angles. Low resolution geometry is a problem but the materials respond nicely to the lights. Feedback is appreciated. Thanks.
Added by:	rashadcarter1
Keywords:	rashadcarter1, bryce6.1, DazStudio
Date:	05.23.2008 05:59
Hits:	4290
Downloads:	68
Rating:	3.50 (2 Vote(s))
File size:	472.3 KB

Previous image: Bouncing Ideas

Next image: 3 Virtual Baked Lights

Author:

Comment:

Render Man
Member

Join Date: 11.10.2007
Comments: 358

I was going to comment on your first try on this scene but decided to comment here. While the lighting looks more natural I still wonder about all the shadows that seem to overlap in both tests. It would be interesting to make this same set up in Carrara to see the difference. Not to compare which program is the best but to see how the lighting and shadows would differ.

I wanted to do some testing but my machine is tied up in a high priority render in Bryce. This looks like it is going to be about a four day render.

05.23.2008 13:05

rashadcarter1
Admin

Join Date: 06.04.2006
Comments: 2610

The bands are due to the number of lights. In a room like this all surfaces, even the floor would bounce light. The floor just beneath the main light is being hit with very bright light. The spotlights on the floor in that area are very bright as they are reflecting alot of light. In the other example you can see on the ball just how much light is shining upward from the brightness on the floor. I used hard shadow spotlights. Had I used soft shadow lights the bands would disappear completely. But then we would be talking about a very long render indeed. That is the problem with the workarounds, they are not nearly as good as a true GI.

I do have a Carrara ver sion and again it looks alot like the Bryce version. I am still working on getting the lights to behave in Carrara but when I finish it I will upload it for comparison.

05.23.2008 16:13

davidbrinnen
Admin

Join Date: 01.03.2004
Comments: 2224

I have no issue with the quality of lighting produced, that looks excellent. I was a bit suprised at the shadows, observed by Render Man - but I understand your reasoning. My main concern is the ammount of time it took to render, which while three hours does not sound much on your machine I wonder how long that would equate to on older less optomised systems. Of course that is not really a valid objection since processing speed increases year on year. So yes, maybe it would be worth using soft shadows in this case and taking the penalty?

05.23.2008 16:58

Horo
Admin

Join Date: 05.26.2004
Comments: 4721

The render time on David's (or my) machine would be 4 or 6 fold at least.

Same remark as for David's Mk IV: It is quite all right to have colour bleeding so strong and shadows so deep. We want to see what's going on. But it is not real. That is what RenderMan says and I concur. I have been looking around my rooms - on the walls and ceiling to find that sort of shadows and colours and I can't make them out. I can see a very faint red projected on a wall near the red curtains and the colour hue is more noticable in the shadowy regions than on the brighter ones. None of my walls get so bright as the one in the room above.

Even though you call this "bouncing light", I still call it reflection. We need a reliable method to measure the amount of light absorbed and reflected as "bounce light". I suspect that we have to deal in a few percents of reflected light, not a few ten percents. We must not forget that a lot of that illumination and light scattering is due to the air particles, probably more that is actually reflected from a plaster wall.

Light always decreases by the square of the distance from its source and Bryce lights to not strictly follow this. And I suspect those of other renderers neither. Comparing a Bryce with a Carrara render will not convince me as far as realness is concerned. Rather, I'd like to see a real room modelled in a raytracing program, lit and compared with a photography. But to do this, we would need a means to measure the light intensities at the source, the walls, everywhere.

Please don't get me wrong, I'm not critisizing your efforts. I'm strictly talking about simmulating reality. Neither you nor David is there yet, but both of you make very noticable progress - while I'm just walking from one dead end to the next. And I'm so grateful that you are going to all this trouble. It made me look at my environment much more closely and concious as I had before. And this is an enrichment.

05.23.2008 21:18

rashadcarter1
Admin

Join Date: 06.04.2006
Comments: 2610

I will conduct more tests.

05.24.2008 08:43

Previous image:
Bouncing Ideas

Next image:
3 Virtual Baked Lights

[Discord Server]