Questions around Renderer for Maya

Become a member of the CGSociety

Connect, Share, and Learn with our Large Growing CG Art Community. It's Free!

Thread Tools Display Modes
  03 March 2018
Questions around Renderer for Maya

Hi everyone,

I'm quite new in the program and currently learning Maya. So far I have seen multiple renderer, including: Mental Ray, Redshift, Octane, V-Ray and Renderman. But I'm still confusing between those renderer. Can you guys help me explaining those advantages and disadvantages of those renderer?

Really hope to receive some suggestions from you guys,
Thank you very much
  03 March 2018
a bit old but a good read... come back if you need more in detail info..

Last edited by oglu : 03 March 2018 at 07:17 AM.
  03 March 2018
Be aware that mental ray has been discontinued. I wouldn't spend time learning it at this point.
  03 March 2018
The article posted has great information but can be a bit daunting overall trying to contextualize all of it in order to glean what is best for you. 

My short version would be to simply get one of the GPU render engines. Octane or Redshift with a huge push here towards Redshift. It is becoming dominant here in L.A. in all applications. 

You want the GPU because on your own you can do quite a bit of freelance rendering on one or two machines with it that was not possilbe before two years ago without a decent render farm or a least billing the client remote renderfarm time separate. 

You should still learn one of the big CPU render engines though as they will be around for a while because GPU hasn't slain all the technical needs that CPU can do, particularly with feature film work. 


Personally I still love Vray the best. I made myself learn 3dsmax just to have access to it and thankfully was quickly ported to XSI and Maya soon after so I could push fx facilities towards it. Which happened at DIgital Domain while I worked there and we helped develop Vray for Maya during the production of Tron Legacy. So I am a bit biased towards it having helped shape its integration with Maya a little here and there. Vray has evolved quite nicely and is hard to beat as a CPU solution. They developed/developing  a hybrid solution where you can use CPU and GPU together though I have not used it. They also created a denoise solution that cleans up gi and shading noise from under sampling allowing you to cut corners with render time in many cases.  Vray has a look and feel that is hard to beat. It has a graceful solution for the most brutal of production demands as we tasked it to death at Digital Domain and Vlado and CO delivered solution after solution with nightly builds. VERY hard to beat the personalization and support Chaos provides. 


Having said that you can not avoid Arnold's presence in the industry. Something I used to avoid like the plague due to everyone being entranced by the preview renderer only to be dragged through serious pain getting rid of noise even after hours of rendering per frame. Somewhere between 2014 and 2016 Arnold became a solid production renderer and I have had to use it on even very recent projects and it does clean up nicely and gets the job done. Arnold still has a less than graceful way to setup certain render passes and material ID's etc. 


Renderman is a bit of a quandry. It was THE way to get things finished back in the 90s. During the SGI era of VFX having Reyes to draw a frame with motion blur and displacement was a kind of miracle. 
Where it had draw backs was that it wasn't really a ray tracer and all reflections/refractions were done with trick maps. It had a brick GI method etc etc. There were Renderman clones that added ray traceing like BMRT (Blue Moon Rendering Tools). Anyways....

This could degrade into a very long useless tangent of information you don't need to know. The point is that the benefits of the old REYES Renderman are long gone as CPU and GPU technology no longer require Renderman style shortcuts. In 2014 Pixar released a new RIS engine for Renderman and that is currently what Renderman is. It has a topdown raytracer approach and GI solution that is closer to Arnold's brute force methods but they have their own GI shortcuts for certain scenes as well. They have also added a noise reduction solution as well. 

Renderman is used by Pixar by default and for many projects at ILM and WETA digital. I do not see it in mass use in Los Angeles. I am of the opinion, having run it through many trials during the RIS beta and every release since that, even though they have very advanced shading and technology that is quite good and makes beautiful renders, it is the slowest of all the big render engines now. Though the technology they have developed for Renderman like the SSS shaders, the hair shaders, and the versatile setups that exist because of the demands of Pixar and Disney are often more a reason for the big studios to use it. It isn't always that they look at render speed. Its usage falls under the ideal of "Is this even technically achievable". All the other render engines have taken/borrowed, even if small in some cases, from Pixar/Renderman siggraph papers as Pixar updates and creates the rendering solutions from scratch. 


This is the render engine I tell people to learn. It is VERY similar to Vray in overall shaders and usage. If you learn Vray or Redshift it is very easy to jump back and forth as the general UI and sampling logic is quite similar. If you can get a 1060 or better GPU you will be hard pressed to find a more satisfying and easy to use render engine than Redshift. Some limitations are due to the volumetrics being limited to VDB volumes. So if you like to create clouds, smoke, fire, etc. you'll need to have a solution which can convert to VDB as Maya native fluids do not draw yet in Redshift natively. One of the few limitations left with Redshift. 
AKA KC / Intuition

Last edited by PerfectLine : 03 March 2018 at 09:53 PM.
  03 March 2018
Very nice summary. Maybe another info can be interesting.

In general you now have two types of renderers. The ones which use more or less pure pathtracing and the other ones which use light and photon caching. The pathtracers like Arnold, Renderman, Octane tend to produce noisy images what is the result of the underlying principle. For that reason they introduces denoising procedures e.g. in Renderman. Vray has it's denoiser as well. The other renderers like Vray or Redshift use a light caching algorithm (or a hybrid version). So first the direct lighting is calculated, and in a second pass the indirect light is calculated. In general the pure pathtracers have less options because you do not have to deal with caches and photon count etc.

Pathtracers have more problems to create caustic effects. That's because they shoot rays from the camera, but caustic effects works best with the reverse way, sending rays from  a light source. In this case we need a bidirectional pathtracer like Renderman which is able to do both. Renderers which deal with photons can produce very nice caustic effects because they simply can shoot photons from any lightsource and use the result to calculate caustic effects.

Do not forget to have a look at the currectly still free Radeon Pro Render from AMD for maya or Appleseed.
__________________ -
  03 March 2018
if i had to decide which render engine we should use for the next animated project i would go with Redshift.... no doubt...
for stills i would go with vray...
  03 March 2018
I’m inclined to take a longer term view. Two years from now I’ll be saying renderman XPU without a doubt

reply share thread

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Society of Digital Artists

Powered by vBulletin
Copyright ©2000 - 2006,
Jelsoft Enterprises Ltd.
Minimize Ads
Forum Jump

All times are GMT. The time now is 04:22 AM.

Powered by vBulletin
Copyright ©2000 - 2018, Jelsoft Enterprises Ltd.