Linear Workflow

if rendering to 8 bit image

  • leave the framebuffer data type at the defalut (RGBA 4x8bit)
  • set the MR Framebuffer Gamma to 0.4545, that way Mental expects all your input textures to have a gamma of 2.2 and corrects accordingly, leaving just the color swatches to deal with…
  • Since you are rendering to an 8bit image, you need to make sure that the image has a 2.2 gamma added at the end. so the best way is to add the lens shader mip_gamma_gain and set the gamma to 2.2

if rendering to 16bit half or 32bit full float

  • the framebuffer data type should be RGBA (Half) 4x16Bit. This is full float, less precision then full 32bit float, but 32bit is a waste of space and much slower to work with.
  • leave the MR framebuffer alone at 0.4545, 8 bit textures that are read in will be linearized automatically.
  • you must turn off "preview convert tiles" in the "preview" section of the MR globals.
  • save to an .exr file in the common render globals. Due to a bug in maya, the image will have .img name.
  • your file will look correct in the render view, and when you render it to disk, you will need to view it in photoshop, nuke, or aftereffects CS3. The viewer will need to add the srgb curve to the image to view it properly. the one's i just mentinoed should add it automaticaly if the file type you've saved is floating point (example : .exr)

You should be using the mia pass shader when trying to go linear

Mia_pass will take care of the color swatches for you. you should be fully linear now.

if you need to go linear when using a shader other than mia_pass or SSS_pass,you must fix the color swatches (use at your own risk!):

  • create a solid color texture
  • us JS_flatcolor or p_constant
  • put that into a gamma node. set the gamma node to .4545 across the board
  • then plut that into the color (or diffuse color) of your shader
Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License