Here is a render that i did after i got inspired from different Disney and Pixar movies.
As this was my first project using renderman for maya, i had used different render engines earlier and i wanted to come up with a workflow to help me populate the environment easily and at the same time optimize the render resources
The goal of this project was
- To scale up environments as much as possible.
- To understand the technical approach to populate environments easily and reduce render costs
- Shading and rendering workflows in renderman using RIS
- Utilizing RIBs to the maximum potential and to study their construction. Utilizing RIBs to seamlessly bridge shading and lighting pipeline by working and updating the library assets and using them to kick off renders from the maya lighting file
- Study and understand the artistic lighting in Pixar and Disney movies and trying to replicate them.
- Using and testing out Megascans textures.
- The entire shot was rendered using a 32 gb ram machine
Modelling and Look development of the assets
The trees and vines were modelled using speedtree for maya and the plants were modelled using both Speedtree and maya paint effects. The rocks were sculpted inside zbrush and textured using Mari using triplanar projection
All the assets were textured using Megascans as i was a BETA tester for their product and I was given permission to use them for my work. Megascans is making a huge library of physically based textures which can scale upto 8k.
The assets were shaded using Renderman RIS shaders
The Center log was scanned using photogrammatry techique and has 8k textures in them
Using RIBs for creating asset library
I used the full potential of RIBs to create a huge library easily. I could export the shaded asset into a RIB which is called into a the lighting scene only while rendering. That helps me to seamlessly edit assets when i keep the lighting file open as the RIB box inside the lighting scene is only a placeholder which calls in the RIB file only when the scene is rendering. I do not have the need to edit any shaders in the scene as thats a very troublesome process.
Layout and populating the scene
The hill cliffs was created using individual chunks of rocks that are placed on top of each other at different scales.
I populated the tree RIBs using particle instancing for maya. So this makes it very light in RAM usage during rendering. I populated the plant RIBs manually using SPPaint3d and it took me just a day to manually layout the smaller plants and ferns.
Lighting and Compositing
I used RIS for rendering the scene. I broke the entire scene into a Foreground and Midground to make lighting and rendering light. I used 3d clouds in the background and rendered them using RIS. I like to render with the least amount of AOVs as i like to keep the compositing cleaner. I like to make sure the final look from rendering is really good at the lookdev and the lighting stage. I rendered the position pass and Zdepth to add extra depth and fog layers. I used Position pass to grade the atmosphere at different areas. I could try this by rendering the atmosphere in 3d with physically accurate gradients but that seemed to be too heavy for my PC so i went for a 2d hack which looks close to the look i approached.
I used RedGiant Looks for final grade of the render.
Overall i had fun learning a lot in this 4 month project. Feel free to write your questions in the comments section below.