So I bought Maxwell renderer. Release date pushed to Dec!

I finally bit the bullet and have bought this fantastic renderer.

No Blender plug in as yet, but hopefully one may be written in the future.

In the mean time I have to model in Blender and export to SketchUp.

Keep an eye out for some of my first renders coming soon.

EDIT - - -

Release date has been pushed to December, but the first screen shots of the stand alone system have been released.

More info here

Some renders made with Maxwell here


Yes, it looks like a very good renderer. It’s also fundamentally different from other ones, as it shows the whole image, and continually refines it, and you can stop it at any time you want. And it’s unbiased, ie: it converges to the phisically correct solution, which means (supposedly) no artifacts.
I’d be very much interested if you could show some test renders, along with rendertimes!

The Renders in the Gallery look awesome.
Never heared of this Renderer… can it render Animations too?

I wish someone would make an open source spectral renderer. I found this link:

so maybe it will come some day. The thing I like about ART is that it supports different colour systems so it will likely be able to produce non-photorealistic images. I’m not sure how well Maxwell would handle that.

The thing that I like about Maxwell the most is being able to choose the render time before starting the render. That would be great for a production environment. Project Orange knows how much time they have to render a given scene so they could push it to the maximum quality possible in that time.

It would be nice if Blender offered a framework that gave access to all its components so that exporting to other formats was easier. Like Maya’s C-based plugin architecture.

Look forward to seeing some renders from you Sonix.

wow, really does look quite good.


One of the Japanese blenderheads, Shizu coded a basic Maxwell exporter with Python and some other tool, from what I remember.

You may want to contact him (though be gentle, his English isn’t that fluent!)

I was thinking about using Maxwell with LightWave. But now I think I’m leaning more towards FPrime.

Sounds cool. I just wonder how this one stacks up to handling large scenes with lot’s of polygons compared to apps like V-ray. I’m hoping that the next re-write of Yafray will have better performance in this area. I tend to render scenes withs lots of complex objects.


From that page he used a Maya exporter after importing Suzanne from Blender. Alas no Python plugin mentioned for Maxwell on that page.

Thanks for the link though.

There are many requests on the Maxwell forums, which are asking for a Blender plugin. I have the Maxwell SDK, but it’s for C++, which I know nothing about. Same with Python I have no knowledge at all.

Is there a difference, or much of a difference between C++ & python?



I have the Maxwell SDK, but it’s for C++, which I know nothing about. Same with Python I have no knowledge at all.

Is there a difference, or much of a difference between C++ & python?[/quote]

They are both object-oriented languages. C++ however is a lower-level compiled language whereas Python is interpreted. One of the advantages of Python is that it is faster/easier to write code with because you don’t have to worry about memory allocation and there is no debugging involved.

But, C++ has a major speed advantage. C++ code can achieve about 10 times the performance of Python and that’s likely why most 3D software plug-in architectures use it (including Maxwell as you said). This speed advantage makes a huge difference when exporting large scenes. Having said that, I’ve used some C++ exporters with Maya and they weren’t all that fast.

The problem with C++ is it’s harder to write good optimal code so in the real world, you might find Python would do just fine. You normally need a lot less code with Python too.

The Maxwell SDK probably won’t be of any use in this case unless by some miracle it’s possible to get anything to read Blender’s binary files so you will likely have to get Blender’s scenes exported to Maxwell compatible format. From what I hear in the pro industry, some people just export geometry like you’ve been doing to OBJ. Then they write shaders manually and use Python to link the shaders to the objects. Then render in another renderer.

It would be nice to have an easy way to do it but Blender doesn’t give you access to all the necessary data through Python. That’s why Renderman exporters have problems. There was a presentation at the Blender conference about it so maybe there will be a rewrite that allows exporters to be written more easily.

You could always make a Python/C module to interface Blender with Maxwell’s C++ SDK.

That wouldn’t be an easy task though.


Slightly OT, but very true. Some creature guys I know work by having a bundle of splines in Maya which they animate, which gets exported to their custom dynamics system, which then goes through a custom procedural modelling system (in their case for feathers, lots of them, all with just geometry) which pumps out a renderman scene, which then produces about 80 million layers for the compositing teams to play with.

Was taking about 3 hours per frame. And most of that was the dynamics, not the rendering.

My first tests with Maxwell render.

Until a Blender plugin is written, I’m going to be using 3DSMax 7 for assigning the materials as SketchUp’s material interface with Maxwell is very limited.


okay, how about this:

I will write a plugin for you, if you give me your serial number.


sorry, I could do it, but I am quite busy at the moment.

pmsl, nice try mate :smiley: I’m in no rush. Hopefully Next Limit will be working on a plugin for Blender, but nothing confirmed yet.


—Tim: awesome renderings…BTW…Would you mind to share how modelling of the ‘full-of-cubes’ wall was made? The results are beautiful.


—Tim: awesome renderings…BTW…Would you mind to share how modelling of the ‘full-of-cubes’ wall was made? The results are beautiful.


Hi, the images made on the Maxwell gallery, used a Greeble script for 3DSMax. I used SketchUp’s Push/pull tool after adding a cube, as I didn’t have the script at the time.

I just kept adding more cubes in as random manner as I could. I then duplicated and rotated this 4 times to make a bigger scene. A similar scene could easily be created with Blender, using AO or Yafray to render. There’s a small amount of DOF as I haven’t got to grips with the whole bluring system yet and this was a quick test. :smiley:

Hope this helps,


Finaly Next Limit have given a release date for Maxwell 1.0, since the 31st October date has passed with no release.

Refinements for release 1.0:-

Huge reduction in noise levels.
Up to 10x faster render times.
No render time increase for complex models.
To be released on 22nd Novemeber.

Up to 10x faster. :o :o :smiley:

Roll on the 22nd. :smiley:


Pretty cool Sonix. Personally I wouldn’t pay that much money for the renderer but those renders are really nice. I’m looking foward to seeing the better and longer renders.

BTW, does anyone here still remember me? The guy who did the StarGate a while back? No? :frowning: