Okay, after countless bouts of testing and tweaking in P5, and previous questions on this topic in the forum....the conclusion I'm coming to is that when you add the glass shaders to a globe prop you will get a perfectly respectable reflection and glare from a light...but not when you want the same effect against a flat plate of glass, such as a store or vehicle window...why is that?

hi pookah, I don't know what shaders you're using, but I can tell you that a full mirror shader works perfectly well on flat panels in poser 5/6 and in daz studio. If a mirror works, I don't see why glass shouldn't. Maybe you could post your settings? mac

I'm using the standard glass shader that comes in P5. I've also tried Ajax's shaders, also to no avail. Where is the full mirror shader you mention above? Maybe I should try that, and play with the transparency?

The miror shader is just one I made myself, but it uses the standard mirror settings that everyone uses. I'll have to look out the settings. Take me 5 minutes. But in the meantime, one crucial question. Do you have raytracing switched on? If not, reflections aren't going to show up. mac

Here are the mirror settings. mac

And here's a pic. mac

Mac, thanks for the settings, I will try that next. In the meantime, here is a rendering, showing the difference between three panels of glass and a globe. All have the same glass setting, but only the globe shows the nice bouncing of light effect that I'd like to see on the panels of glass. I have three spotlights pointing at each of the three panels.

And yes, I do have raytracing on.

Mac, how come when I plug the reflection node into Alternate Diffuse, as you do above, I don't get background, quality, softness and ray bias settings, but only background, quality and softness?

Hmmm.. could just be a curvature problem? Or rather, lack of curvature. mac

Those glass panes, how are they made? Planes or boxes? That may make a difference. If you make a glass pane from a flattened box, you'll have both an entry and exit surface. For nice refraction effects you'll have to set the raytrace bounces to at least 4. You won't get good refraction effects when you use a single sided or double sided square. The glares in real-world flat glass are influenced by slight imperfections on the surface. A tiny little bit of displacement, driven by an fBm shader, might help to get you those glares. I'd say displacement setting at the equivalent of 0.1 mm (0.004 inches). By the way, if you have Poser 6, you should use the fresnel shader. It combines reflection and refraction nicely - I've used it to make some very convincing water. Hope this helps, Steven.

I think sdvl is right about the displacement. With no curve to catch the light, you're going to have to 'rough up' the surface. I did the mirror in P6. Maybe that's why it has RayBias? mac

We're really talking about specular hightlights here, not raytraced reflection. Flat planes will return hightlights but what tends to happen is that the whole plane gets lit up by the highlight, not just a little hot spot like what you get on curved surfaces. If you fiddle enough with spotlights or environment reflection maps you can probably get something that looks OK, but it's not easy.

Yes, the conclusion I've come to, after a full day of fiddling, is that it is best to Photoshop a reflective surface on the flat plate of glass. Thanks though, for helping me learn.

Another element of the problem is where the reflections come from. Can be done with reflection maps, or maybe an ambient light texture on a scenery-flat behind the camera. There's a set of free MATs from RDNA which apply a tree-branch pattern to lights, good for woodland scenes, and the method might be applicable here. But it is all down to where the reflections come from.

1. With lights pointed directly at the reflective surface, there'll be some washout in the entire effect - so fiddling with the lights is crucial. 2) My experience with raytraced planes says the fewer polygons on the plane the better. Raytracers tend to calculate per polygon which can easily result in a fragmented image. (see 3 below) 3) The model origin needs to be considered. If the model came in with smoothing, and many objs. - particular those exported from P4 - will contain smoothing data. To help prevent fragmentation, make certain the plane is not smoothed. Absolutely critical if the plane has more than one face. - Mark

Why is the reflection node being plugged into the alt_diffuse channel in the sample image above? Why not the reflection color channel?

After giving up on trying to render the right effect in P5, I photoshopped this. I'm happy with the result, which used photoshop's glass block filter, and lens flare. Any thoughts?

Lens flares won't work in P5, but you could try using a spotlight with a wide angle and a short distance end, and volumetric lighting. I don't know how that'll show up in a raytraced reflection though. I do know that render times will be long - very long.