Lately I have started picking up Vray. My initial thoughts are that it’s hard for me to draw a direct comparison between it and Mental Ray as I left 3DSMax around the same time Mental Ray was being added to it, and so I have never used Vray in Maya and never used Mental Ray in Max.
That said, I am so far enjoying my Vray experience. Everything seems relatively intuative, and tends do do as it says on the label so far which is nice.
One thing I am impressed with is the simplicity of its render pass system (actually, called ‘render elements’ in Vray). So without further ado I think its a good idea to get these notes about reconstructing a render from its various elements down here while everything is still fresh in my mind.
First things first… you need to be working with a Linear Color Space the whole time since many of these elements such as Raw GI will be linear. (In the coming days I’ll put up some information about setting up for linear workflow with 3DSMax and Vray).
- Also, before you start rendering out your passes, you have to define where we want to save the files: Open the Render Setup dialog and browse to the V-Ray tab. Open up the V-Ray Frame buffer dropdown.
- Enable ‘render to V-Ray raw image file’.
- Click the Browse… button and find/create a folder to save the image sequence in.
- Switch the save as type to OpenEXR (*.exr) and hit Save.
So firstly the render elements required in a basic form are:
- Raw Lighting
- Raw Global Illumination (GI)
- Refractiion (If needed)
The basic order in which all this gets comped is as follows:
- Multiply the RawGI by the Diffuse.
- Multiply the RawLighting by the Diffuse.
The results of these two comps would be the same as if you had rendered a GI element and a DirectLighting element respectively.
- Add the two resulting layers from above together.
- If you used refraction, add it in now.
- Add the Reflection element on top next.
- Add the Specular element on top next.
At this stage, your resulting image should resemble just what you would have got from your renderer if you had just saved it out as a simple rgb image without any render elements. The benefit of using elements is that you can colour correct or adjust each of those elements individually in the comp without having to rerender from 3D. Combined with the floating point, linear workflow, this provides an amazing ability to control the look of the final image (and eat a heap of disk space).
Additional elements, such as SSS passes and so forth can also be added into the mix, its just a matter of finding the right place for them in the stack and setting them to their correct blending method (add/plus or multiply).