What makes a RENDERER great


Lets start a thread on what makes a renderer great.
Now we kind of started a thread over here on tail end of Turtle for Maya.
We decided it was going of topic so I want to start a new one here.
Anyway you can read what we said here, thread starts at 41

I want to keep this constructive as possible, I want to know why people think Mental Ray is so powerful, most people no its programmable, you can right your own shaders, edit MI files etc…
But is there anything straight out of the box that makes Mental Ray brilliant, things that regular artist can use functions/features that you don’t need a TD for ?
I know a lot of people will turn there nose up at this but, I do mainly Architectural Visualisation and I find using FG/GI all these new rendering techniques brilliant. So I find Vray a dream, the irradiance maps produce great detail !!

So start with the other post, but add any comments to this one and tell me what you girls/guys think.


You have to start with the idea that a renderer only needs to be great at doing YOUR projects (not every project) for you to consider it great. For example, some renderers are considered great because they can render high-geometry, high-resolution scenes with displacement mapping, motion blur, and depth of field, quickly. That kind of “greatness” is based on the needs of specific types of companies. Other people might care about other things: how easy is it to use for an independent artist, how well does it take advantage of hardware rendering capabilities of your graphics card, what kind of global illumination functions are built-in? Other people might consider a renderer great if it comes at a low price, and is open-source and easy for a programmer to modify. All relativism aside, I tend to think of greatness as coming with the different projects, and that even if the reputation rubs off on the software, it’s the work that 3D artists create with the software that really establish what’s great or not.



So its really coming down to project specific and personal opinions.
Is there anything that you personally like about Mental Ray ?
I’m just starting to use XSI, so I’m trying find out what people like about it.


i guess speed is a common denominator no matter the final results. I think there are also little things in renderers that are there but people might overlook sometimes.

I work on some commercials and a GREAT antialias is always required, so if your renderer gives you good AA cappabilities thats an option that cant be overlooked. ALso the posibility for compositing channels. Brazil R/S had a nice iniciative in doing this, but it went kind of in the wrong direction i guess, i mean there needs to be a renderer that you can tell you want an especific thing as a sepaparate channel,but im not talking about the common things like diffuse, color specular, etc… i talk about gettin in every single aspect of what composes a rendered image and no trying to fake the channel you want wich can be very time consuming.

Also a renderer that adapts to other effects plugins you might use is important, check out how many problems people has sometimes with shag hair and Gi or stuff like that.

Also something that people often seek no matter the final product are Motion Blur and DOF, this too guys can make your live a living hell in production if you made a bad choise, it is incredible on how few renderers manage true motion blur and fast depth of field.


Programmability is a must for film work, as well as rock-solid performance, procedural geometry generation, super-fast antialiasing, displacements and motion blur…


the question is: what makes a picture great ?!

the renderer is just a tool.


Right now I’d settle for depth maps that don’t flicker (I’m looking at you, Maya).


No, the question was, what makes a renderer great?

Sure the renderer is just a tool… but if it can’t finish that sequence in time to get it to comp, it doesn’t matter how good the picture is, you’re still screwed.


I found a thread that has a lot of information about the renderers. As I red it, I found that PRman would be the best in the matters of quality and speed. It has the best and fastest motion blur and DOF. The problem is ease of use. Mental Ray is also pretty good but when it comes to motion blur it is a living hell.

Here is the thread (quite long, but very informative!)


there are a couple of things I like about Mental Ray,the first one is the AA which gives me pretty quick good results,another one is the FG,that provides more realistic feeling to my images ,even though FG can sometimes be very unnatural in some scenes (you see…it always depends on the scene you have)…but because I’m a maya user i can’t really and honestly tell you Mr is the best of all the renderers,'cause it was the only choice for me…Last Night I installed Turtle…I’ve begane to practice with it…it’s quite simpler than Mr but I don’t feel it would be my choice at any time…i’m waiting to see V-ray…:shrug:

one thing I know for sure about Mental Ray is that the DepthOfFiled takes an entire life to be done!


speed, flexibility, stability, quality…
something like speed and stability is easier to measure than say flexibility and quality…heh heh.

like the idea for the thread though…but as jeremy points out its about needs of the user…however does every user always know what they want. I know it could be argued they should but very often it would be left to researchers to specify/define the need into something useable in an application.

dunno its a difficult one to answer…maybe you should concentrate on defining what makes a bad renderer! then you’ll know what makes a great renderer…heh heh.:shrug:


Just some things that tend to go into what i think about in terms of what makes a renderer good:

Quality- Sort of speak for itself. But more specifically be able to solve all AA and filtering
problems. Both actual sampling and cheats play in here…

Reliability- When you feed it certain settings, how reliable will it be in terms of output
quality. Will it render 100% without problems. Will it render 5 frames with black splotches or
something similar…

Memory management- How much can you feed it it terms of bitmaps (mb and pixel size), polygons and whatnot, coupled with a bit of how it treats displacements

Raytracing (in all forms) - Maybe one of those ‘spoiled’ things, but can solve a lot of problems you might get into without it…

Lights, luminance space treatment- Does it support arealights/shadows, how does it work with shadow maps, does it clamp the output?

Programability/Scripting- How much can i change to make things better for me, either by new
‘technology’ or just making a UI work better and so forth.

Cheats- Need to support all kind of cheats you might want to throw at it.

Shaders- How flexible is it? How reliable? How cumbersome to create ‘new’ ones?

Motion blur- In some form or another so it can output frames with smooth motion blur, sharp
details that respects opacities and raytraced thingies :slight_smile:

Quality flexibility- Not only how much I can push the setting upwards, but how fast can you get it for those speedy testrenders…

Anyway, just my 2c :slight_smile:


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.