---
title: "Rendering Basics"
canonical: "https://rmanwiki-27.pixar.com/space/RU/629571614/Rendering%20Basics"
format: markdown
---
Ever wondered what happens under the RenderMan hood when you press the render button?

Well … here is a simplified explanation of the process.



1. Read scene description into the renderer (or update it if already in)
2. Tessellate (and optionally displace) the geometry; build ray tracing data structures
3. Generate and shoot camera rays
4. Shade ray hit points: evaluate shading networks, evaluate materials, compute direct illumination
5. Generate rays for indirect illumination; trace those rays
6. Go to step 4.  Repeat.





  
Step 1 Scene translation  
Step 2 Path tracing  
Step 3. Display the data  
Step 4. Denoising




**Scene Translation**

Terssalates geometry / makes micro polygons

Puts data in RAM (system or/and vram)


TIP: Keeping gemotryt as polygons is cheaper, but renderman prefers sub

fact: you never need to chosse dsub d leveles, renderman does it all for you, always smooth


**Path tracing**

shooting rays from camera

hits surfaces, 

shaders discovered and rendered


TIP: Procedural shading networks will always be more expensive to render than textures.  If you really procedural networks, you can bake them to disc.


**Display the data**

Display the render (IT or viewport) or save to disc (2D image or Deep data)




**Denoising**

Clean up path traced varience and make clean