[Michael Betke]

Some client told me after he wanted a change to a reflection of an rendering:
"At the end its just a simulation of reality. No matter if your tools use physically correct math."

He was right. hehe

[JDHill]

I disagree.
In real world you don't have hidden emitters and AGS...
So maxwell cannot be totally accurate anyway !

Okay, let's try this again. If you model a building upside-down, floating a meter above the ground, Maxwell can render that accurately. It could only be called unrealistic in the sense that nobody knows how to build an upside down levitating building. AGS, hidden emitters, hide to GI -- these things are no different; they are rendered accurately by the engine, and are only unrealistic in that we do not know how to construct such things (yet), in our physical world. Bias would be if you gave the engine a model of a room, and it rendered one corner over-bright due to the use of an algorithm which used a shortcut, a bad assumption, or a deliberately incorrect model of how light behaves. Do you see the difference?

[choo-chee]

hmmm nicely said but I think we are getting too philosophic here ....
just give use less noise in less time. somehow.

[dmeyer]

Clearly any engine that does not simulate the totality of the nuclear fusion reactions generating energy and thus light to power their "physical sun" engine is biased.

[hatts]

Why not mess around a bit with FIRE?

No one expects to use FIRE for their final render, so why not make it even more useful as a crazy-quick preview? I think a large majority of users would be fine with FIRE using light caching or something. If the user already knew that FIRE wasn't producing a 100% correct solution, they would know when to use FIRE and when to use a full render.

Also, good one dmeyer

[JDHill]

That suggestion seems to miss the whole point of Maxwell Fire. It can be used (e.g. in the free SketchUp plugin), but is not primarily intended, for rendering final images. It is definitely not faster than Maxwell Render, it just picks at the low-hanging fruit first, in order to provide more useful visual information (you don't exactly need to resolve 23 refractions to confirm whether you've misplaced a light in your scene), quicker. But that comes with the cost that it can take far longer to resolve complex lighting. All that aside, the real point is that it does not render something different from what Maxwell does. What you talk about above is an entirely different thing -- a different engine, using biased algorithms, and that is going to appear in neither Maxwell nor Maxwell Fire, because it would completely destroy the purpose of either. If you need a biased engine, you should use a biased engine...tools are tools, and you shouldn't expect your screwdriver to perform anything but poorly as a hammer. Maxwell as Swiss-Army Knife has never made any sense to me, as other solutions exist, and as any effort put into such a direction must necessarily come at the expense of development of the engine -- which is what people are asking for, in threads such as this one, when they request that Maxwell be made to render faster.

[rickyinmotion]

in conclusion, gpu rendering is not for maxwell

[JDHill]

If you're responding to me, please point to the place where I discussed GPU.

[JTB]

Fire is a great tool... but I have never been able to make it work in MAX... it crashes a lot... probably this is a MAX's fault (one of the ten millions).

[raduc]

Fire is a great tool... but I have never been able to make it work in MAX... it crashes a lot... probably this is a MAX's fault (one of the ten millions).

It shouldn't crash that often, can you send me a sample scene you're having trouble with ?

[choo-chee]

I don't think it's MAX only since I use Maya and before clicking "fire" I ALWAYS click "save"...
tons of crashes, esp. with displacement and stuff imported from CAD software.

[l1407]

maxwell redner can't use gpu, do you change from path base redner to point base render

[sandykoufax]

With Max2013 and newest maxwell plugin, I'm very happy with Maxwell Fire. Any crash occured so far.