Bryce5.com

Registered users

» Forgot password
» Registration

Random image

First bryce picture
Comments: 5
luciavds

16.04.2024, 07:24

Home Page

Google ads below

Portal light 1

Portal light 1
Description:	I confess, I've been playing mindless video games - portal. And the first thing that struck me was how good the lighting was in the environment. So, naturally, I began to ponder if it could be replicated in bryce. The trick being, to get a global lighitng effect inside an enclosed space - without the programming means to bounce the light arround (like Carrara can for example). The room and the boxes were wingsed up. The room has a window, a door and a slot around the edge of the ceiling to let in light. Top left. Lit with Horo's Jeeprobe at maximum quality. Top right. Lit with Jeeprobe at quality 64 and soft shadows preimum effect 64 rpp. Bottom left. A none shadow casting light above, seven dim lights in the scene and a single bright light at the window. Bottom right. Back to the Jeeprobe. The room set to not cast shadows. Then a faded box placed around the scene. And a fuzzy sphere placed above the ceiling to create softer shadow regions within the room without the penalty of preimum effects on the render time. Again a bright light was placed at the window. Something to ponder.
Added by:	davidbrinnen
Keywords:	davidbrinnen, bryce6.1, Horo, hdr, jeeprobe, lighting, experiments, portal
Date:	10.26.2007 08:34
Hits:	3885
Downloads:	82
Rating:	0.00 (0 Vote(s))
File size:	202.0 KB

Previous image: Protal light 2

Next image: Wolf test

Author:

Comment:

Nightst4r
Admin

Join Date: 12.11.2003
Comments: 451

Portals!

I swear I had the concept and basic implementation for portals 5 years ago... Even been working on it recently.. till i saw they released it already and now I'll Look like a rip off.

=(

10.26.2007 13:35

Horo
Admin

Join Date: 05.26.2004
Comments: 4721

It's not the first time I have the impression that we tackled the same thing at the same time. As long as we don't know what you were up to, it is difficult to point out the best choice. As my mind is set currently, I think lower left is the best, it is the most neutral in shadows and colour. Your experiments are very interesting. I'm very surprised at the result from the best quality, there are ray-patterns on the walls which I think are difficult to account for at 4096 lights. But it might come from the grey-white pattern of the probe used, which is essentially MoonBlueGrey.

I think all four methods have their merit and it really depends on what you need for your scene. The results are very different and your tests are enlighting. Thank you for showing this to us.

10.26.2007 18:18

rashadcarter1
Admin

Join Date: 06.04.2006
Comments: 2610

Hmm. I guess this is as good of a time as any to find out what the heck is going on with the soft shadows. The top right example is rendered with just 64 light samples, yet you say soft shadows are enabled. Exactly which soft shadows are you talking about? Does setting soft shadows in the sky lab affect the hdri/ibl shadows? I have not observed the two settings interacting with one another, I have tested it many times. I thought the only way to soften ibl shadows was to set a higher number of samples. If you were to tell me that the top right example was both jeepprobe AND radials or default sun set to soft shadows, then the soft shadows would make sense to me. Please explain. It's almost like your version of bryce works differently than mine. Perhaps there is a setting in the render options I am not aware of. Please let me in on the secret. Nice work, interesting test!

10.26.2007 21:16

davidbrinnen
Admin

Join Date: 01.03.2004
Comments: 2224

That's what I like about Bryce, just when you think you've been over everything a dozen times or more, some new facet reveals itself. So... what happens is this, if you engage Premium effect soft shadows while using a light probe for lighting, soft shadows are applied to the simulated light sources of the probe. I think credit for discovering this goes to Dan Whiteside - but I could be wrong there... anyway, I'm sorry, I thought this topic had already been covered, didn't mean to be keeping you in the dark. As for the banding on the top left image, yes Horo, I was dissapointed by that, I suspect it is due to the narrowness of the aperature through which the light is passing - which is indeed very narrow.

10.26.2007 21:33

rashadcarter1
Admin

Join Date: 06.04.2006
Comments: 2610

Aha! Premium effect soft shadows is the only soft shadows that works for the ibl. I did not know. It is not your fault, I just didn't know myself, as my computer is far too slow to engage premium effects most of the time. A little further experiemntation and it would have revealed itself to me. Thanks for the explanation. Based upon the render examples above I would say that 64 lights at soft shadows probably looks 1000x better than normall AA at 4096. How do the render times compare?

10.27.2007 15:33

davidbrinnen
Admin

Join Date: 01.03.2004
Comments: 2224

The upper two renders took about 12 hours each the lower pair took about an hour. The pay off for the premium effect soft shadow is the impact on the speed. Turning the effect on was worth over 4000 simulated light sources in terms of render time. Which is why I dropped the quality. I wouldn't like to guess at how long 4096 light sources would take to render with soft shadows, but I would suggest that it would be worth while looking for another apporach to the problem. Horo gave me an idea with his last image. But using blurred reflection (controlled by speculative halo) and metalicy to try to hide the reflectiveness. It is a trick I've tried before but unlike the sky, when the hdr is rendered as a background, it can be made very bright indeed, even if the monitor can't reflect this, the renderer responds accordingly. So I'm trying to reflect very bright background hdr images into the scene to light it up. No need to use the hdr light sources, just a single source as Horo did in the scene and a very bright background which can be seen through the "lights" in my room and the "window". Anyhow that's the theory. The render times are cruely long so it will be a while before I know if it's any good as a theory.

10.27.2007 16:43

Horo
Admin

Join Date: 05.26.2004
Comments: 4721

There is one information missing: the number of rays. Except for the sample with the soft shadows engaged, we see steps in the tile joints, something that would disappear if rendered larger or using premium render with more rays, 16, 64, ... So the render times are not completely comparable.

10.27.2007 16:49

gat
Member

Join Date: 12.21.2006
Comments: 667

wow, I have no idea how you did the lighting, but the last two look awesome!

the one on the bottom left has the best shadow distribution, but the one on top right has the best AA

10.27.2007 17:36

davidbrinnen
Admin

Join Date: 01.03.2004
Comments: 2224

Horo is right about the rays, the only one rendered with premium was the top right (to get the soft shadows) and as a bonus, better joins between the tiles. For ease of memory, top right was quality 64 rpp 64.

Gat. Well, the one on the bottom left is a relitavely easy one to set up. It was Rashad's use of non shadow casting secondary lights that prompted me to put in a light above - this has the advantage of creating some "ambient" light that can get more or less anywhere, this light was set to squared fall off. Otherwise, seven dim standard fall off shadow casting radial lights were placed in an arc over the camera and a single stronger light in the middle of the "window". IBL I wanted to use for this subject, because well, it's the new thing. But IBL does not entirely suit interior scenes because it is tricky to get the light inside. IBL is fine for a still life, since the "room" then becomes the world. Or for outdoors, it's great - so with typical awkwardness, I want to do interiors.

10.27.2007 18:59

rashadcarter1
Admin

Join Date: 06.04.2006
Comments: 2610

Food for thought. IBl for indoor scenes is a sure riddle, takes nothing short of genius to solve. I am very pleased with the results of this test and the lesson I have learned. Good stuff. Thanks a ton.

10.27.2007 22:17

gat
Member

Join Date: 12.21.2006
Comments: 667

I see........... so it wasn't magic! :D

10.27.2007 23:31

tina gazcon
Member

Join Date: 08.07.2006
Comments: 254

Very interesting experiment. They all look totally different. You are truely an expert in lighting David. I am likeing the two on the right. Very nice!

10.30.2007 14:31

spektyr
Member

Join Date: 07.02.2005
Comments: 1010

AH! You do such interesting lighting jobs, David. I sense you like playing with lighting like I like playing with reflections. ;-)

11.03.2007 01:31

Previous image:
Protal light 2

Next image:
Wolf test

[Discord Server]