Maxwell Render | Official public forum

First image: Using only skydome.
Bounces=5
Sampling level reached=14
Rendertime = 60min



Second image: Using emitters (planes) on the windows
Bounces=5
Sampling level reached=16
Rendertime=6hours



Maxwell needs a good deal of speed optimization ...

That's the first time I've seen that clock rendered without it looking like it's floating in space. That's what I love love love about Maxwell. I'm a secondary shadowing junkie I guess. I hope there's a 12-step program for that...

Oh, and cool circuit board.

clean
metal city is a cool idea, very conceptual....

Frances, can you believe it took me two-three weeks trying to do this with AIR and the results were not as good (but it would render in 2min with simple occlusion)?
On the other hand it took only one day figuring out the Maxwell settings about one hour figuring out whether to use emitter or skydome; and 6 hours to render.

Tom, thank you for the nice comment. As you can see we do not have texture maps or bump-maps available yet... so we can play with only simple things for now (which is good; it is better to start with the basics )

-

IMO it doesn't look "right" with big emitters at the windows. It doesn't look natural, instead it looks artificially lit, which in this case I guess it is So it's right, but it's not "right"......it looks a bit flat.

Would be interesting if you continue this render until the caustics from the glass table clear up. I suspect x3 rendertime. This is why the suggestion to have Maxwell focus on areas which are still noisy is terrific. I mean if you can render a region in much less time than the full image, why not automate this?

Which gives the idea that you can use bitmap region rendering and only render that portion, then composite them together.....so this already works in a sense....am I writing this down?......but it would be nice to have a workflow like: Stop the render when you see a part that is satisfactory, exclude that section with a marquee selection, continue rendering.

This is an unbiased renderer, and like daros proved, there is no problem splitting a render into regions. Biased renderers can't do this since they have to calculate first the photon or light map for the whole image and only then start calculating the regions. Maxwell doesn't have this limitation so I think it's important to take advantage of it fully

This is why the suggestion to have Maxwell focus on areas which are still noisy is terrific..

Mihai, I aggree with you 100%. As a matter of fact I had similar thoughts earlier today.... some surfaces clear up so fast and they are completely clean within level 13... if only Maxwell had a clever mechanism to detect and exclude them

IMO it doesn't look "right" with big emitters at the windows. It doesn't look natural, instead it looks artificially lit, which in this case I guess it is Smile So it's right, but it's not "right"......it looks a bit flat.

Isn't that awesome ? I put emitters in the windows and the room looks exactly as it would if there were emitters in the windows (I couldn't fool Maxwell !)
So you are right, it looks more like an excibit room in a museum or something (having recessed lit areas instead of windows).

(after some more experience with the settings), I will have to understand the tonesmapping controls and how to succesfully throw light from outside without bleeching the windows and without making it look flat... given some time ... I am working on it

Nice render, Thomas!

What is the material parameter for the buildings in metalcity and tv screen and metalic light cover in the second image?

tetsuMaxwell wrote:Nice render, Thomas!

What is the material parameter for the buildings in metalcity and tv screen and metalic light cover in the second image?

tetsuMaxwell,

The lamp in the room was aluminum with UV rougness 0.3
The buildings in the city were silver with UV rougness 0.25

-

Thanks for your quick reply Thomas!
It was simpler than I thought..

How about the material for the tv screen in the room?
I like the vague dispersion on reflectoin.

tetsuMaxwell wrote:Thanks for your quick reply Thomas!
It was simpler than I thought..

How about the material for the tv screen in the room?
I like the vague dispersion on reflectoin.

Same easy thing.
It was plastic with very very dark gray color and specular and with UV rougness 0.3

Thank you for quick reply again!

I have difficulties on defining the balance or diffence between diffuse and specular colors in plastic material.

What do you think is the best way to define them according to the amount of reflections?

The values of color and specular need to be very close to each other (almost identical).

OK. thanks, Thomas.

I appreciate your kindness.