Vray or Mental ray, for photo realistic characters
Hey guys, first of all, this is not going to be vs thread, just i want a option which one to choose, for rendering. For both Vray and Mental ray, i can't find some tutorials about it. every tutorial i found is for creating interior and exterior, but nothing for achieve realism of character.
Some help/advice/suggestion?
Some help/advice/suggestion?
Replies
The Vray docs are all on chaosgroup's website, under the "support" tab
Rendering characters that are both light and dark skinned, in different engines? o.O
Again the render is just the tool if you learn one good enough you won't have any problems.
Vray has that softer lighting, and SSS from what I've seen, and it captures a lot of the more subtle nuances of the more translucent caucasian skin.
Blacks have darker more opaque skin that lends itself to sharper contrasts and less scattering.
I'm not racist! I swear! I hate you all equally!!!!
For example, VRay comes with alot 'established' variables that make it easier to control right off the bat, but is extremely unforgiving if you don't have the right maps and put in the wrong ones, opposite of Mental Ray, in which an artist will have a hard time gripping with certain points, but you could use the same Diffuse map for all slots in SSS, and you'll still get decent results.
Also, in favor of MR, I always come across 'Uber' shader for MR created by the community, in the case of SSS, I found an Uber-SSS shader that had stuff like Oren Nayar, secondary 'Bleed' values and etc integrated in it, something which I THINK (not sure, since memory is fuzzy) that VRay might not have many of.
Also i want to know about this 3 point lighting? How to set up, which lights to use it?
I see some good render from the people in their portfolio, does they sue this 3 point lighting thing?
Whatever gets your project done.
(I use Maya and MR because that's what I learned in school.)
Vray is what most studios are moving to. Mental ray is old and inefficient. Stuff like depth of field and sss will render faster and better in vray. Vray's lights are easier to control IMO. If you are trying to simulate a real environment you control the light temp with a degree of Kelvin, not a color in the color picker. Also the lights are based off geo, which makes it more like real world lighting. It is very easy to take a piece of geo and convert it into a light. For example if you are trying to replicate a ring light, you simple convert a donut.
Another thing I like about vray is that you can convert the camera's into "Vray Physical Camera's". This allows you to have all the controls of a real camera.
As far as SSS goes, yes the main example is skin, but it goes beyond that. It's used for liquids and candles. Anything that light passes through and dynamically changes the color.
Reflections are also treated with a more realistic behavior. Sometimes that gets in the way of your "vision", but realize if that's the case, your vision won't look real.
And last note, do not use vray sun and sky. It's a noob move.
If you have any vray questions feel free to PM me.
Photometric lights have offered kelvin controls in Mental ray for a long time now, and making light emitting meshes is as simple as applying a material.
so far as camera simulation goes i presume your referring to exposure control, MR offers mr photographic which mimics films reaction to light. gives fstop and iso controls and so on.
What's so wrong with sun and sky in vray? I know some people just whack a default rig into their scene without proper consideration but a light is a light providing its physically plausible then its up to the user to expose it right and create good materials.
I don't know where kelvin is in mental ray but it is an easy switch for every light in vray.
I guess mental ray has common camera settings but I like how vray handles them more.
Sun and Sky is what people use when they don't know how to light. Yes, once I used the create sun only and angled it but you can build the same thing from scratch and usually you will get better results. I'm just saying the usually when someone brings up sun and sky, it means they are too lazy to create their own setup or they don't know how.
Another thing to add is that the vray_mtl is a lot more powerful and uses better termanology in the settings than the default maya shaders like blinn, lambert, and phong.
I guess the decision between mental ray and vray is whatever like to use, but more and more companies are moving away from Mental Ray and going towards Vray. Also Arnold is starting to become more popular.
I prefer VRay to every other renderer so far.
Here's a monkey:
Here are also the render settings I use most of the time. I know it is maya but it should cross over I think.
Thanks for posting your basic settings Talbot.
Just like to ask since i haven't used vray for maya yet, can you use the native mental ray shaders in maya and render them out in vray?
One major thing to note is that I used to use Maya, and at the time I was testing the two Vray was much more adapted to Max. I tried Vray on both and there were some definite differences. So if you use Maya I would be a little more wary, MR and Maya work easily together and I've had a meh experience in the past with Vray and Maya. If you use Max then Vray seems like an easy choice (if it's a choice), some of the best renders I've seen used Vray in Max.
cgfeedback has some good examples, Vimmy is the first that comes to mind http://www.cgfeedback.com/cgfeedback/showthread.php?t=414 (nsfw)
Edit:
He describes the shader setup he uses in the thread
IIRC, someone actually wrote an Uber-Skin shader for Mental-Ray for Max, and it had all the important functions in, even when VRay doesn't (Oren Nayar anyone?), not to mention has mutiple layers of Specular (Gollum from LOTR had essentially 3 Blinn and 3 Phong calls) while VRay has one at most.
At the same time, VRay and MR cannot hold a candle next to Maxwell, especially when it comes to the more 'organic' shaders, like skin, hair, etc. especially in terms of sub-variety, in which they have several specialized one you can mix and match through blends, as opposed to 'tick box on separate mesh' approach of VRay or single mesh material in MR.
for the single artist without much rendering experience who wants to render turntables or moderately complex scenes v-ray offers you a sturdy set of training wheels to keep you from falling over and cracking you head on the curb. setting up shaders and lights is fairly easy because everything has been tuned to give you a 'nice' result out of the box. the render setting are also refined to the the essentials. in most cases you have to adjust 2 or 3 setting to get a final image.
mental ray has all the features that v-ray has and a good deal more. but for the single artist without much rending experience you will never want or need to touch any of those extra features. mental ray has no training wheels and therefore offers you many opportunities to crack your head open and fall into pits filled with snakes etc... there are situations where you may need to wire shader parameters by hand, construct custom shaders from nodes or even write custom shaders in c++ to get certain effects.
if you stick to the basics, area lights, HDR dome lights then your ok in either one. and you can get essentially the same image in both renderers. each one has speed advantages in certain effects. if your looking for a set of training wheels for mental ray then take a look at mentalcore http://core-cg.com/ . it gives you a more v-ray like setup.
hahaha , you made my day sir!
go for the one you know more, both give outstanding results on good hands.