[deflix]

its a different algorithm - not progressive refinement - its loose incorporation into VIS was a nod towards existing LS users nothing more. Autodesk bought lightscape and killed it for one reason only: it was too ahead of its time, and wouldnt allow for a slowly advancing and therefore profitable buisiness model. Take a look at this message posted on a editing forum some time ago. We should really boycott this company somehow....



This is the **** that ran it till 2005 carol bartz



And this is the **** thats been ceo since.... Carl Bass



Just so everyone knows who is responsible for ruining such a beautiful and special industry.

[Frances]

I don't have any of my Lightscape work to show, but I did these with Viz 4 Radiosity in 2003/2004:

http://www.surrealstructures.com/images ... y07_03.jpg (link)


http://www.surrealstructures.com/images ... y03_03.jpg (link)


Stumbling blocks for me with radiosity: poor bump-mapping, no caustics, no SSS, no glossy reflection (except for the LS workaround of a very fine noise in the bump channel, which didn't work with Max/VIZ because of the poor quality of its ray tracer).

It's kind of weird to open old Max/Viz files and see the radiosity solution in the viewport. I stopped using radiosity in late 2004 and used Vray for a couple of years.

[gianca]

I do remember Lightscape, good stuff...
Ten years ago I imagined how powerful Lightscape would be today... Back then I would never imagine SGI will be relegated to a boutique firm and AD would swallow most of the existing CG IP landscape...

There were obviously limitations with Lightscape, which was very efficient with flat surfaces but it showed artifacts easily on curved surfaces. And of course, as you mention intersecting planes are ALWAYS an issue with any baking solution I have knowledge of.

Baking obviously can't do reflections and refractions, so raytracing have to be added after the diffuse calculation, and that wont easily take in account the interaction between all these elements, especially with Maxwell layers and coatings.

Still, I do see a value in using unbiased renderings for Matte Paintings: in a unbiased workflow once a rendering get approved, the issue is how to extract the elements that can be used in a baked renders, and how to replicate as close as possible the additional passes that cannot be precalculated. If that's not challenging I don't know what it is.
We are probably going to use Vray, since it's speed and VFX pipeline capabilities still outweigh any benefit of an unbiased render...

But it doesn't have to be.
I hope I don't have to wait another ten years...

Gianca

lightscape was a very good rendering software with two single issues: It was very complicated to build correct 3d models because "progressive refinement" algorithm didn't' liked intersecting planes. Lightscape wasn't very suitable for organic shapes rendering.
By side that it had a great but simple material model, powerful lighting tools ( the best ones i ever sow) Incredible baking capabilities (with geometry to texture functions), rock stable and easy to learn.
All baking technologies i know have the same problems as lightscape... intersecting planes and arganic surfaces.

[def4d]

Thanks, it's more than interesting, i must have a look!

[deflix]

I don't have any of my Lightscape work to show, but I did these with Viz 4 Radiosity in 2003/2004:

really great work! - always been a fan.....

[def4d]

Frances, i think i saw one of your radiosity render in ~2003, i must say i've never saw such beautiful results that yours with that engine

[Bubbaloo]

Pity, I can't see the pictures Frances posted.

Edit: Nevermind, I see them now. And they are very good!

[Frances]

Thanks guys.

Brian, sorry about othe images. Sometimes my web host doesn't show them inline right away on forums. I hope it isn't interfering with my web site though. Most of the time nowadays, I use my Cox Communications web space to host images on forums to avoid red x's.