Techniques for that "Pixar" effect in Mental Ray 3.5


Hi guys,

Can anyone give any tips/techniques for the “Pixar” style of rendering in Mental Ray?

It seems to be a mixture of cartoon and photorealistic, and I love it. I can’t really justify $1k for Renderman for Maya at the moment though (and the latest version doesn’t yet support Maya 8).

Is it mostly procedural textures/shaders as opposed to texture maps?

Many thanks,



It has mostly to do with lighting, not textures. Lots and lots of lights to get a softer look, and plenty of rim lights, which helps give it a slightly more graphic look. Why not post an image of your favorite shot and maybe people here can give some suggestions on how to achieve the look.

  • Neil


Hi Neil,

Thanks for the reply.

Could you just clear up “rim lights” for me? I’m a lottle lost :slight_smile:

Some of my favourite images are from Nemo (Bruce the shark’s skin texture, that kind of thing), but I rarely have the need to personally do underwater shots…

I’ve attached a couple of shots from Cars, which to my mind is amazingly produced.

  • The Route 66 shot is photoreal (to me), but at the same time you can tell it’s a cartoon.
  • The interior shot - the wood texture is superb, and although it’s stylised, the shot again looks almost photoreal. I love the rusted texture on the pickup as well.
  • dirty.jpg - I would love to develop a shader which would allow you to “add” dirt and dust to a surface.

A generalised thing about Cars as well - the car bodies are superb, with the Porsche’s slightly matted look, and the hero’s bright red gloss, (and of course the pickup’s rust).

Neil, no you’ve mentioned lighting, I can see that the scenes are often much brighter than “real life”.

I just love the hybrid real and toon look that Pixar films have :slight_smile:




Just found an article on cars on this site:

Interesting info about using area lights.



Well first off, I did not light on this film, so don’t take this as gospel, but here’s a few tips that may help you get something similar.

First off here’s a really useful lighting discussion that has some great ideas, especially if you’re new to lighting…

Lets check your images now, check the first image below for what a rimlight is. Basically, one rule many people follow when lighting is always have a black surface on a white surface on a black surface on a white surface, etc. The idea is to setup your objects into visual layers so that the viewer can easily tell where one object stops and another begins. Rim lighting is one way to help do that. So in the night time image, say you have a darker McQueen against a dark background. Well, the eye won’t be able to immediately tell where McQueen ends and the background begins. So you want to create light and dark planes. In this case, you light the right side of mqueen to be light (probably a fill light, or your key), then the left side of him goes dark, then right at the edge where the darkside of mqueen ends and the dark background begins, place a rim light, which is just a bright light placed such that all you see is a little sliver or edge of light on the corner of an object, so that your image now reads light (right side of mcqueen), dark (left side of mcqueen), light (edge of mcqueen), dark (background). This layering approach is one thing that helps contribute to that slightly cartoony look.

Notice the same pattern in the background of light against dark, see my image number 2.

This also helps define surfaces, say you simplify everything in the scene into cubes. If you’re seeing 2 or 3 sides of a cube, one side should be darker then the other, because then the eye will read the shape of the object. If both sides of the cube are almost the same brightness, then your eye won’t be able to tell what the shape is you’re looking at, and everything will appear flat (which might be what you want in some cases, but not to achieve the look you’re specifically after.) See my image #3 below.

As for the outdoors daytime scene, you can get something similar in any package with a skylight (which will create nice soft shadows), and a keylight that represents the sun. Also notice the use of atmosphere, as things get further they disappear slowly into fog, which gives the scene scale, and notice how the shadow areas go purple, an artistic choice which I think worked beautifully.

And the courtroom scene, all I can advise is a lot of soft lights :slight_smile:

  • Neil


Oh, and in that article you pointed out, Bill Cone talks about Maynard Dixon, check out an example of his work (beautiful stuff, BTW)

And compare that to your outdoors in the sun image.

  • Neil


One last post, I promise. For even more info, check out this book:

Chapter 13 has a really good discussion about lighting from Sharon Calahan, who was the Director of photography (lighting) for Finding Nemo.

  • Neil


Neil - thank you! Those notated shots are fantastic. Things are starting to be a lot clearer. The layering approach is something I will be studying in detail.

I have ordered the Advanced RenderMan book from Amazon.

I think, for my next project, I’m going down a Max route rather than a Maya one. Still haven’t decided yet though. I’m sure I’ll end up doing some things in Maya, but I need to get my Max knowledge up to date, and I recently got 9.

Incidentally. I’m applying for a Pixar internship (if I’m eligible) next summer (BTW, before you think I’m being nepotistic, that’s NOT why I’m asking about the “Pixar” look :smiley: - this new stuff won’t make it into my showreel for another year or so!)




pixar use radiance lighting with renderman, this is a global illumination technique to have environment colored shadow. In mental ray you could have the same result but you have to use photon, final gather… pixar use a new technic of GI without raytracing than anyother rendering engine could achieve at this time, so it cost a lot of time in mental ray to have the same result. after all is a question of talent and lighting art. all rendering at pixar is prepainted by pastel artist by hand, so you have to think a lot like a painter to give your atmosphere to the final render. so use ambient occlusion pass added to the GI pass with color bleeding for shadow coloration and some good direct and environment light and map with proprer shadow. pixar mix raytrace with deep shadow. of course the best is to use renderman because mental ray GI is not as in control than the renderman one.


Cool. Just remember, having specific pieces of software are certainly a bonus in getting a similar look, however, achieving a look is far more about the artistic choices you make than specific software, especially since the feature set of most major renderers and 3d applications are pretty similar these days.

I mean, take a look at the short “Gopher Broke” by Blur Studio, many people commented on how it has a pixarish look, and yet it was all rendered in Brazil (with maybe a touch of scanline renderer here and there).

So don’t worry about the software you use, worry about the style and visual language the artists have used, and then find a way to take the parts you like and change it into something new, and something that’s your own style.

  • Neil


Hi Neil,

Yes don’t worry - I wasn’t doing the “I use cool software so my images are going to look amazing automatically” thing there :smiley: . It was just a passing reference…

Max 9 and Maya 8 use the same version of Mental Ray anyway.




of course cgi and software are always just tools to enhanced your creativity, keep in mind that the most beautifull part of your final image come from your imagination and talent to recreate that. this is why the best software will be nothing with the good artist to drive it.

the only difference in rendering is the speed and render time. For sure with renderman this effect is faster because pixar devellop renderman for themself first and for their movie, so it’s a big part of the quality but as I told you you could do the same with mental ray with a little bit of practice. You should be a little bit slower in rendertime than with renderman but you could do great thing with. elsewhere you have to wait the next Rfm 2.0 wich will be out in one month.


Do you have any related information on this topic ? This sounds a little doubtful. is it already implemented in PRman 13 ?
what i heard is that indirect ilumination in prman shall be rather slow although i have to admit that im not very familiar with it.


its called point based oclusion and colorbleding. Its in renderman 13 but not it Rat yet but in studio pro and rat 7 it will be more ready i think…


Could you elaborate a little bit on the point based occlusion thing? Is it the same as mental ray’s final gather. Irradiance caching? Brute force?


Im not shure how they do it actually Its a mapgen procedure almost like subsurface scatering in renderman but for oclusion and bleeding. I have not tried the new point oclusion in renderman only what i have heared on the pixar forum… I have to wait for RAT 7.


It’s a new technic to compute color bleeding and occlusion and indirect illumination with the light information of the scene. pixar use a file system to store information of illumination in the scene called point cloud, the point based occlusion use algorythm to create an approximation of this point cloud data to create GI, color bleeding and occlusion.
they used it on “cars” and it’s really fast, approx. 3x faster than raytrace and much more with displacement. this technic is cool because it’s really smooth GI in the result for animation, no noise, no flickering at all. we save hours in rendering now with this technic and I’m sure pixar will imrpove a lot this technic in the next release. it’s only for renderman 13 it will be in Rfm 2 and renderman studio.

point cloud could also do subsurface scattering as meshman told , it’s a really powerfull technic.


This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.