Build 302 and nXt engine 4

[George Ioannidis]

Completed render test using new algorithm. Wall bump pops up really nice, the glass is very convincing, chrome sphere does some really nice caustics on the teapot. The only thing I don't understand is the stripes visible on tea cups, especially on the nearest one. Looks like surface softness was lost somewhere.

[Roy Hirshkowitz]

Nice.  I'll have a new build pretty soon.

[Richcat]

Hi George - luv it - but couldn't resist giving it a quick clean to get rid of that dust - it was far too life like. Wish we could have a go with the sketchup version.

[Roy Hirshkowitz]

I'll let Al have it pretty soon Richcat-- just a few more details to work out. 

[Richcat]

Cheers Roy :-)

[Roy Hirshkowitz]

With sun and caustics.

[Jorge Javier Lara Domínguez]

Es sol automatico Roy? Se ve muy bien la imagen

[Roy Hirshkowitz]

Just "Exterior Daylight" preset.  No caustics switches to set.  No glass in the window in George's model, although I have a version with glass which I'll test soon (should be a very similar result.)

[Roy Hirshkowitz]

Sun (slightly different angle) and single sheet of window glass.  ~15 minutes-- taking a little longer to resolve but well on its way.

[Roy Hirshkowitz]

Standard glass with IOR of 1.5. Because this one uses a glass window, all of the light in the scene is really a "caustic" and is calculated quite differently from the first image.  It is encouraging that the results are similar.

[Roy Hirshkowitz]

Fresnel for glass is based completely on the IOR-- the fresnel slider on the Transparency page should always be left alone.  Same image with window glass modeled as two 3dfaces.

[Roy Hirshkowitz]

Ah-- nothing at all to do with fresnel.  This light path is excluded in all of the engines (intentionally).  You cannot see the sun directly, nor can you see a perfectly specular reflection of it (you can't see it through a perfectly specular window either.)  If you turn on Highlight under the Sun settings you will get some reflection of the sun in materials which are glossy (not perfectly specular)-- but I don't generally recommend this.

I have considered allowing this path and may do so in the future-- but it's difficult to handle.  The solar disk is very very much brighter than anything else in the image.  The brightest part of this image is about 25000 cd/m^2.  The solar disk can be as bright as 1,600,000,000 cd/m^2.  It's tough to say what it actually looks like because you can't look at it comfortably.  It would also cause an enormous firefly problem...