[Peder]

It is my second interiror rendering project where this has happened to me. I setup my render in FormZ and when I go to render in maxwell it looks like the sun is not there. Last project I freaked out and spent hours trying to find out what was wrong. Eventually I just let the render run and lo and behold finally the sun made it's way into the image. The glass is AGS type so there shouldn't be a refraction penalty right?

As you can see from this image at SL 9.28 the sunlight is there although looking suspiciously weak in the preview:
http://idisk.me.com/peder.lindbom/Publi ... 081542.jpg

And here is my OGL image in FormZ:
http://idisk.me.com/peder.lindbom/Publi ... 082102.jpg

Any ideas?

[Bubbaloo]

Try turning off the glass geometry to see if its an AGS material setting that's causing it. You may even find that it works better with no glass in the roof. The only reason to use AGS in interior renders is if you want some interior reflections on the glass surface. Remember that the more reflective you make the AGS, the less sunlight will get through.

[tom]

Even, sun caustics from real glass would show up in SL2. It could be the specular part of concrete material revealing them late. Besides, SL9 is not high enough for specular surfaces. What happens when you render using quarter resolution?

[Peder]

Yes I am trying without AGS right now. It looks like the AGS may be reducing the intensity of the sun -that may be the cause. And yes the reason to use AGS is that I was hoping for some reflections.

It seems this has changed since 1.7 somehow. A colleague of mine remarked that he had seen something similar. And Tom if your suspicion is correct shouldn't the sun show up on the wall in the background which has a diffuse material?

Could there be some conflict with using 1.7 materials in a 2.0 scene causing issues? Could using thin SSS leaves on a tree that has other for the bark but still is made into one single object by Onyx cause issues?

Or is this simply a case of get a stronger machine or let it run longer?

It seems counter intuitive to the basic concept of an unbiased renderer to have the sun illumination appear later than the rest of the illumination. In that sense it is only unbiased after a definite render time... No?

This is withouth AGS and on SL 6. It looks like a similar issue in this one although the sun looks brighter in the preview window:
http://idisk.me.com/peder.lindbom/Publi ... 150102.jpg

[Bubbaloo]

Try testing it at a lower resolution. You're rendering at 2500 px, at S.L. 6 the lighting will not be "filled in" yet. Let a 800 px render for a while and see how it looks all clean.

[Peder]

Thanks for the help. I don't have time for more trials atm. I have to hope the sun shows up eventually. I have put the renders on a stronger PC so hopefully I will get some sunlight in the end. It is strange however. I did some preliminary render tests early on that cooked for a few hours only and they don't show this effect. My feeling is that there is something else going on either in the geometry added or the materials. Here is a noisy example:
http://idisk.me.com/peder.lindbom/Publi ... 155431.jpg

[Peder]

Thanks for your efforts to help guys. It turns out it was a simple case of unbalanced emitters and no multilight. I didn't pin this as a cause because I thought the sky looked similar to the other renders. But as it turns out when running the render to a high SL using Rebus and enabling Multilight. Sure enough the sunlight was there all along.

Sorry for being an i-diot.

[Bubbaloo]

Glad you got it figured out!

And you're not an idiot.

[Peder]

Well. Guess I spoke too soon. No there is something going on here. The sun does come but it comes in much later than the rest of the lighting. Here is a screendump from another view. As you can see the sun is starting to make it's appearance but only after 6hrs of rendering. And if I am to judge by the render I made on Rebus it won't look good until much later.
http://idisk.me.com/peder.lindbom/Publi ... 083847.jpg

Can't believe I am the only one experiencing this?

I am on a Mac by the way but I did try the same views on a Vista 64 bit machine with Maxwell 2.02.

[Peder]

Here is the first render with the sun developed after a lot more rendering on Rebus.
http://idisk.me.com/peder.lindbom/Publi ... 084929.jpg

The floor material is -I believe-provided with Maxwell. It's called WoodenFloor.

[Mihai]

What are your emitter settings? Could you do a test hiding all the ceiling emitters to see if sunlight appears faster then?

[Peder]

Bingo!
Mihai your instincts were right. Here is a screenshot showing the material settings for the lights, the lights themselves inside FZ and a quick test render at SL 6
http://idisk.me.com/peder.lindbom/Publi ... 144122.jpg

The question now is why?

[Bubbaloo]

How many emitters are there? I see the wattage you set at 400,000 watts! Try to stick to real world values for best results. 100-200 watts max. per emitter x # of emitters.

[Peder]

OK, I didn't know that this mattered. There are 415 emitter objects consisting of a flat square polygon with the normals pointing down. If I assume the lighting consists of 2 fluorescents each at about 35-50W then the emitter should be in the range of 20-40K watts. I made another test with the emitter set to 20k watts. And it didn't make a change for the better. It only looks a lot more noisy:
http://idisk.me.com/peder.lindbom/Publi ... 150835.jpg

[Asmithey]

Are you using the latest Maxwell plug-in for Form-Z? If you want, I can take a look at it for you and test it on my end. You would have to laod your files on Divshare or something so I could download them.