Vray for Maya on Arch Viz VS Mental Ray (Stills/Animation)

[cgbeige]

Comparing render times with two different renderers is kind of an apples/oranges thing. And I don't know if anyone here has the same file set up as final for both renders. I bought VRay for Maya after years of being frustrated with Mental Ray, even though I only used it as an illustration style "semi-realistic" renderer and Maxwell for my money shot photorealistic renderer. Compared to Mental Ray Vray is noticeably faster for displacement and motion blur renders (motion blur in MR is retardedly slow), it works with instances and GI with no hitches and it consumes far less memory. Most importantly, the render passes and linear workflow just work, which was what I wanted in an illustration renderer. I don't do animation but I see around here that Mental Ray seems to still have issues with flicker. It's not clear to me what MR actually does right...

It takes a lot more work to get close to what Maxwell gets by default in VRay or Mental Ray. It's just the nature of the Maxwell renderer - it's slow because it does everything by default and avoids the photon vs light intensity garbage in MR which is inherently physically incorrect. How can you have a light that emits X amount of light of this colour but its photons are completely different and you're left guessing what the right balance/exponents are? Then make another light and solve the problem again - in MR, your images aren't getting more realistic, they're just "less wrong."

Maxwell was written much later though so it's a completely different way of solving the problems that are very clearly incrementally solved in Mental Ray and to a lesser extent, VRay (research paper comes out, FG goes in; research paper comes out, AO goes in, etc). But Maxwell's realism is also its weakness - there are no optimizations like light cache blending, sample overrides, etc so animation with it is uselessly slow. You also can't turn off camera DoF since it's a camera simulator. You also can't do volumetric things like smoke with Maxwell and it's render passes are very limited. But it is the best realistic renderer you can get and after three months of using it, I was getting images with one light source that only MR technical directors get with a ton of faking lights, post work and separate AO passes, etc. There are talented people out there who use Mental Ray to get stellar images but Maxwell is just like turning the light on and taking a picture - the rest is just tuning scene scale and shaders, which are similarly extremely simple but physically accurate.

[grynds]

What i meant about comparing render times was using the same project with mr lights and mr shaders and then with vray lights and vray shaders.

Anyone did this?

I think that would be useful to compare both render quality and render times.

I've never tried out Maxwell but from what i've read it's just like you've said, good results but really heavy render times and not very good for animations.

cgbeige, you've got some nice Maxwell concept renders though.

[anevsky]

Good breakdown cgbeige!

The only thing I could add is that I look at these render engines like tools in a toolbox. You don't have to stop using the screwdriver forever because youv'e now switched to the hammer. You're still in the same base application so you can work with each renderer when it is right to do so. I have often used different renderers for different passes or aspects of the same shot even.

[cgbeige]

ya, I agree. I have a comp in this project that is a combination of 3 renderers since it was just the most efficient way of handling the shot. If I ever learn how to write MEL, I'd like to write a script to do a light rig transfer setup for MR > VRay, MR > Maxwell but that is far off.

[CGTalk Moderation]

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.