Why Vray and not Mental ray ?


Seeing as MR comes free with Max and it is a recognised pro renderer used as standard in many high end apps, i wonder why so many people choose vray when they have MR

What exactly does vray offer that MR does not have ?


I’m not saying this is necessarily true but I’ve heard people say that Vray is easier to learn. I’ve never really tried to learn mental ray a great deal so I can’t speak from experience here but I did find vray fairly easy to pick up as did workmates of mine


A couple of months ago I tested the VRay demo quite considerably to see what all the fuzz was about.

I tested it until my demo license ran out and I went back to Mental Ray right away. :slight_smile:

My reasons where:

  • VRay, in my opinion is much more cluttered, unorganized and harder to learn and use than Mental Ray.
  • In those scenes I tested it it performed worse and looked worse than Mental Ray. Of course people will say I didn’t use the right settings, but I tried a lot of “best” settings found on the web and fine tuned quite a few of my own and still I preferred what MR gave me.
  • The VRay material felt messy and unintuitive for me compared to the A&D material in MR.
  • No matter how I tweaked the different GI methods in VRay I still think Final Gather is faster and gives better results.

What I did like about VRay:

  • The frame buffer is really awesome for tweaking stuff
  • VRay Displacements were very, very fast and used almost no memory at all
  • The physical camera is great
  • vray-materials.de is an awesome resource

But as most people say, every renderer can produce good results as long as you know how to use it.

I guess it’s hard to avoid VRay if you venture into ArchViz seeing how established it is, but for me, who’s just a 3d hobbyist, Mental Ray is the way to go.

Also, an added bonus when using Mental Ray for ArchViz stuff is that people tend to say: “That looks very good for being rendered in Mental Ray”, as if MR can’t produce good results. :slight_smile:


I can tell you my reason - Vray is faster.

Every time I’ve attempted to light/texture/render a scene in MR, I get fairly decent results but the minute I try to render at a production resolution (4’000+) MR dies a horrible death.

I’ll also admit that this could be due to complete ignorance on my part… But I think there’s a reason you often see high-res print (like architectural) work being done with Vray.


An obvious pro, though not necessarily unique to VRay is 3rd-party plugin compatibility.

Although it’s starting to be addressed with the news that FumeFX is going to be mr compatible soon.




the answer is simple…

all renderers do only what you say to him. When you are good, is the renderer good. Learn rendering technology and you can use this renderer what you want. It is wrong that vray is easy to learn and the other renderer not. When you make a good picture is vray the same as other renderers, you must have the knowledge and time.

hot chip


I don’t exactly agree, I come from a MR background myself and I like to use it for my personal projects BUT, this renderer is in no way “at least its implementation in max” is good for a studio - farm envo.
-RPC and FG and Backburner, deadly mix, vray is solid.
-Rendering any big arch scene with MR above 2k image res is a risk, you go for a coffee or a break and come back and find some “Black buckets” in your image, vray:vrimg.
-it’s pretty much incompatible with any other plugins whatsoever.
-sometimes you get violet-red buckets in your image, and you spend half an hour looking it
up, to find a half-solution to it which leads me to:
if you read mr reference, it’s just a maze of physics and stuff that are completely irrelevant to your deadline, vray has a clear and crisp explanation of everything AND some lovely warnings (If you tick this box, you might encounter this), which mr lacks them.
Not to mention until recently mr had no proxy system, and that is just un-usable for big projects with a lot of trees etc…


I think both (or every renderer) have their adventages.
As far as I know it’s much easier to achieve good results in Vray. But this has its consequenses, which mainly means less flexibilty. Although vray comes with the SDK it’s pretty messy and poorly documented, while Mental Ray is much more opened with its shaders that you can get from web or write by yourself (this, on the other hand, means you have to learn/read a lot). Two of vray’s biggest adventages are good displacements and relatively fast true 3d motion blur.
From my knowledge it has always been real pain to get nice motion blur in mental ray fast.
Morover although it comes free with max with unlimited number of render licenses, the connection and implementation is quite poor and limits the possibilities.
All in all, there’s probably no renderer that is simply the best and also it’s certainly a matter of the specific industry you are working in.


Having had a few years experience with each renderer, I’ve come to like and dislike different things in each one:


*Lighting and GI is superb and under the right hands looks just as good as the unnbiased renderers out there, only faster.
*VrayProxy (from my experience works much better and stable than mrProxy).
*Very well documented.
*Very good VFB, better overall memory usage, has already been detailed in this thread.
*Compatible with most plugins.

*Quite a steep learning curve, the parameters are endless and it’s a real skill to get scenes to render fast and look good too.
*Not free or cheap.

Mental Ray:

*Versatile shaders.
*Easy to learn.
*Very fast and reliable for simple lighting conditions.
*SSS, AO shaders are very fast and reliable, better than Vrays.

*GI & FG are not great and generally just really ‘primitive’. It’s only recently even possible to get a decent interior render, and still not worth the trouble.
*Incompatible with just about anything.
*Documentation is rare and looks like it’s geared towards programmers and not users, you usually have to translate things from German forums to get any useful information.
*Memory problems wherever you go.

In conclusion, if I were looking for crsip realistic interiors, such as those needed in Archvis, there’s no doubt I would choose Vray for my studio.
If I were doing character animation I would probably stay with MR, as it comes free with max.
But as always, it’s always best to know both :slight_smile:


Does photon implemtation mean that MR copes with camera gi aminations better ? Can a gi photon map be done in one pass that renders the whole animation ?

In vray the lightcache is calculated once for the whole animation, is something similar available with photon mapping in MR ?



Vray and mental ray have no problems with GI + Animation. I am only a Artist and not a seller.:slight_smile:

hot chip


Photon mapping in mental ray is not sufficient on it’s own, you need FG with it for a decent image, but yes you can use one pmap for entire animation as long as nothing is changing except for camera xyz, Vray has a photon map by the way too.
I have to say that photon mapping especially in mr is very buggy, if being expert in mr photon mapping = being able to tell if the orange dreaded “No Photons Stored after emitting blah photons from blah light”, you can never be an expert, sometimes it’s double-faces, sometimes normals, sometimes samples, but for the most part, for no apparent reason at all; mr seriously needs something like an irradiance map…
Lightcache can be calculated for entire animation rendering only the first frame, but it’s really impractical for long ones since it’s gets so huge that it max crashes.
The most common approach for animating with Vray is to:
-Baking Lightcache from 0-x using 24 samples per frame or more depending on how “changing” your animation is, so your samples are like 4096.
-Baking Irradiance map incrementally every 2 frames.
-using them to render the final animation.
In mr:
-Calculating photonmap for one frame, BUT if anything is moving especially light you have to RECALCULATE it again ALL OF IT, no incremental stuff, for each frame, sometimes this is a disaster.
-Activating FG, you have to chunk your file to 0-200 frames chunks, otherwise FG will just tick off and tick on again on a certain frame (very ugly), and then “google” how some guy used command prompt to merge the FG maps together, very user friendly huh?


Come on, I know it, you know it, we know it, mental ray is behind vray, in terms of image quality, support, features, etc. Even autodesk says this. And vray’s future looks bright with soon a 1.5 SP3 and 2.0 somewhere in 2009, so it is a good investment in terms of skill.

Why do you also think that there are a lot of people willing to pay 800$ to replace something that allready comes for free with the package. Mental ray is great, it evolves too, but bottom line is that it lags behind.


Hi raymarcher,

come on, you know, you have not the knowledge about Rendering. One Question: Why is mental ray the most using Renderer on the World and not Vray. So many Movie are making with mental ray, why not Vray? :wink: Because Vray is for small Projects good, like a Indoor Scene, but not more. I work for Animation.com (the Number one in Germany for Architectural Viz.) and we use Vray only for Indoor (low Budged) and not for the Big Projects.

A lot of poeple pay 800$, because you don´t need Network Renderlicense. For mental Ray (stand alone) must you pay for each Rendernode.

hot chip


Come on, I know it, you know it, we know it, mental ray is behind vray, in terms of image quality, support, features, etc. Even autodesk says this. And vray’s future looks bright with soon a 1.5 SP3 and 2.0 somewhere in 2009, so it is a good investment in terms of skill.

Why do you also think that there are a lot of people willing to pay 800$ to replace something that allready comes for free with the package. Mental ray is great, it evolves too, but bottom line is that it lags behind.

This is an uninformed statement. We use both renderers here, and have for years, and your statement is simply not true.



Quoted for 100% agreement. And yes we use both of these engines and more. :wink:



well both renderer has its own pros and cons…both renderer can give great image quality if used correctly… Personally I found mentalray very effective and faster and new BSP2 in max 2009 is rocking… I am using it a lot with Backburner in mentalray without any problem and rendering goes very fast on quadcore.

I am using mentalray a lot in my VFX production pipeline and for rendering more than 10 million of polygones(ofcourse with xp64bit and 64bit max) and it renders like a charm.

In my last project, I had to render heavily Material displacement on few objects (which were using Gold metal A&D material) and thanks to BSP2, it was very fast and very stable.(final polygones count went up to 27 million polygones with complex objects…no kidding)

I personally prefer mentalray(efficient and very powerfull) becoz of it comes with max and no need to buy extra renderer plugin.

Mostly you need to use FG for better GI solution.

In the beginning most people found vray very good compare to Mentalray maybe becoz of it wasn’t implemented in max very good and was hard to understand (also bad log. exposure control) whereas vray, final renderer has many exposure controls available and few GI options. so with available exposure controls, vray rendering looked very good than MR. but now things are changed since max9, exposure controls are recorrected and implemented well and as well as new MR features.

at last it depends on user how he uses any renderer. Every renderer(like vray,mentalray, FR, Maxwell) can give great quality if used correctly.


that is was i mean, use that what you need. We use XSI+3dsmax and that is the only reason for use mental ray in the most Projects on my company.

Sorry for my last answer, it comes from my heart not from my brain. :smiley:

hot chip


Its all down to what you know better. Since I loaded max 200964 I found Mray a dream to use, it had been opimised and tweaked to make it far more user friendly.
Having said that, I have delved into vray more, just in the last week for a project with a 7 storey structural glass satircase. And where maentalray falls over and runs out of memory, vray deals with the project. I have always found vray to be betyter with memory managment. And recently, understanding how irradiance works and how to tweak it properly, Ive cut my render times almost in half.
It is true that vray has its settings which affect the render quality spread wide, but once you learn that lights, aa, irradiance settings and DMC sampling affect different specifics, you can usually deal with most problems.
For me they keep flipping, but mainly I come down more on the vray side of things.


It seems the Mental Ray implementation in Maya and XSI is quite superior to MR in Max.
It works with VFX tools like liquids, smoke, fire, etc. in those programs. Not so in Max.
Also, Maya 2008-2009 has a dandy IPR, and XSI has a very fast render region (auto updates with every change) tool. Max has neither.
Both finalRender and VRay have IPR’s (Interactive Preview Render) in beta, and due to be released anytime now. Each of these two have some features that you won’t find in MR (like the Adaptive Quasi-Monte Carlo GI engine with flicker-free modes for both Camera fly-throughs and Animation, Dynamic MSP, and finalToon) , and they are just about bulletproof regarding compatibility with Max plugins. MR will give you fits in that regard. You’re always finding something that it doesn’t work with or doesn’t like. That’s the chief reason I bought a seat of finalRender. Until Autodesk and Mental Images get that sorted out and offer a Max implementation that’s on par with with Maya and XSI’s, I’d highly recommend giving them a look. Brazil is another alternative.


FR Stage1 R3 IPR video: