[Bf-blender-npr] Blender Extended Expressive Rendering (BEER) overview

[Light BWK]

Hi,

Thank you Ton for starting this mailing list.

Blender Extended Expressive Rendering (BEER) is a speedy way to create
shaders for npr.  It breaks away from the common workflows. It is a
result of careful consideration and observation of many rendering
engines.  The BEER proposal is a result of over 30 blenderheads' ideas
and solutions.

Link: http://wiki.blender.org/index.php/User:LightBWK

In brief BEER is about:
1. Layered shader system: because most shader equations are
layer-able; presented as layers, folders and group, with generic
styles as preset.
  a. Object shaders.
  b. Screen space shaders: realtime post effects.
  c. World space shader: shaders which interact with all object shaders.
2. Realtime viewport preview: shaders are animated most of the time,
machinima workflow.
3. No nodes, but there is code layer, either shaderFX/cgFX/pyFX/etc.
4. Pre-composite workflow: no need to wait for render to see shaders
composited, only heavy composite effects will be done in compositor.
5. Fix old BI limitations.
6. Provide framework for future NPR advancement.

Or the longer version written in August 2013:
http://blendernpr.org/beer-a-wordy-leet-intro-august-2013/

BEER would be at the cutting edge of NPR, breaking from past
workflows.  It is to have among the most simple and yet productive
workflow, which most of the proposal artists agree on.

What we hope to solve is loosely like a metaphor of building a house:
1. Identify code foundation as corner stones for development, must be
flexible for future expansion.
2. Partitioning each room according to function, so that it will be
clear where to develop what.
3. Building the walls, doors, windows of each room and its content.

The first step is to understand what is written and shown in the
proposal from an artist's POV.  I'm willing to help in the process and
get more related artists to give developers a broader perspective.



Best regards,

Bong Wee Kwong (LightBWK)
representing BlenderNPR (BNPR) team