[Peder]

Thanks. I have PMed you the links.

[Asmithey]

Peder I joined all your emitter objects into one object. I got sun right away with no errors and a nice bench mark. Try selecting and joining your emitter at let me know what happens.

Aaron

[Peder]

Yes!
Success at last. Is this a confirmed bug then? It has cost me a few more gray hairs, a couple of lost nights sleep and 120 Euros worth of renderfarm time...

Hope it helps someone else at least. Sigh... And thank you so much Aaron for tracking it down...

[Asmithey]

Actually it is not a bug. It is always a good rule of thumb to group and join your emitters especially when you have so many. It takes a lot of ram and resources to calculate all the emitter individually. Glad i could help.

Aaron

[Peder]

No I disagree on that point. The whole purpose behind Maxwell is that it is supposed to be unbiased. If the emitters don't develop simultaneously how can you call it unbiased? OK maybe I made a mistake by not joining the emitters. But my understanding was that this was no longer necessary with V.2 since the multilight function was calculated based on materials instead of objects.

And still I am not completely convinced this is solving the problem. Now I get sunlight but do you get any light from the emitters? I don't see that light (yet). And if i boost their strength I get the purple preview window that I mentioned in another thread but didn't get any reply to.

http://idisk.me.com/peder.lindbom/Publi ... 032518.jpg

http://idisk.me.com/peder.lindbom/Publi ... 032702.jpg

Anyhow, I'm getting grumpy. Better go to bed now before the (real) sun comes up again...

Thanks again, I really appreciate your efforts.

[Asmithey]

I forgot that emitters work differently in 2.0.

If I really crank up the emitters to ridiculous wattage with the multi-light slider, my preview turns pink as well. This looking buggy to me as well. So I am getting the exact results as you. No emitters sun shows immediately and with the emitters on the sun does not show until late in the rendering. This must be a bug.

I think you should send these files to someone at Pylon to see if they can help. It may be with the pulg-in?

Aaron

[Peder]

I will be more than happy to provide the files to anyone at NL just let me know who...

My hunch (I could be wrong) is that this is not plugin related...

/Peder

[kami]

If you join 10 emitters, you have to multiply the value for that emitter by 10. That may be the reason why they are so weak now.
And there is this bug which causes the preview to turn pink at very high voltages, but your image should look fine anyway.

[Fernando Tella]

In MW2 you have to multiply the wattage only if the lights are the same mesh. MW distributes the intensity over the surface of each mesh, so each mesh gets the intensity specified by the mxm. You were giving each light 20000W as they were different meshes.
You got an indoor sun

[Peder]

OK but I still don't understand this. I did try joining the geometry -essentially creating a single mesh and then the light was very weak although it was still 20k Watts. Shouldnt that show up more? And when I boost the intensity I get purple noise in the preview.

And in the other case where the emitters were not joined the light showed up fine from them. It was just that the sunlight looked like it was coming through glass and creating refraction noise. There was nothing wrong with the intensity of the skylighting otherwise.

In what way does the above relate to what you are suggesting?

[Fernando Tella]

Hmmm, hmmm, you are sure the scene is to scale with correct units, right?

[Mihai]

With the separate emitters, you applied a 40k watts emitter to each one, which is an incredibly strong emitter With this much power, the optimizations in Maxwell won't work very well and that's why you see the sunlight later in the render. Best is to keep the emitter strength to normal values. With V2, if you use the same emitter material for all those separate emitters, you'll only get one ML slider which saves you memory, so it's no longer necessary to merge geometry first. If you did merge the geometry then you would have to multiply the emitter strengh by the number of emitters you merged. But in your case, this is not necessary.

If the emitters look too weak with a normal emitter value, then that's how strong they will look compared to the amount of light coming from a bright sun lit environment. So it's best to keep this balance to a natural level both in terms of realism and render time.

[Peder]

OK. Mihai I took your advice and lowered the emitter strength to 100W. Still with separate objects for the emitters right? I do not see an improvement. This is all academic for me at this point so I don't know if it's worth pursuing. I am just hoping this discussion will benefit someone else with similar problems. I provide the MXS file for you to dissect at your leisure and lay down my pen now.

http://web.me.com/peder.lindbom/AIX/Maxwellbugtest.html

Screenshot here:
http://idisk.me.com/peder.lindbom/Publi ... 182901.jpg

Thank you, good night and good luck!

[Asmithey]

This definitely needs looked into because am getting the same results and can find no way to make it render properly. I believe there is no reason the sun light should be effected this way when emitters are on.

[Bubbaloo]

Lighting can act strangely when motion blur is on. Any chance it got activated somehow?