Skip to content

lwander/open-km

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Kubelka-Munk in GLSL

This repository contains a working implementation of Kubelka-Munk theory, but with fake K/S data. This being released as an educational guide, and to help those interested in developing their own imaginary pigments.

Background

Kubelka-Munk theory models the color of a mixture of pigments. To do so, it uses information about the pigments in the form of their K (absorption) and S (scattering) curves. These curves relate wavelengths of light (typically in the visible spectrum) to the amount of light absorped or scattered.

Kubelka-Munk theory is both effective and efficient when it comes to simulating pigment mixtures. However, it is difficult to acquire the needed K & S curves to perform this mixing accurately. Doing so by hand requires a lot of careful data collection and computation (see this thesis for more information). In addition, the K & S curves are considered to be the IP of the paint manufacturer, so they are not keen to have them shared once known.

This repository contains an implementation of Kubelka-Munk pigment mixing, but with fake spectral data for the supplied WHITE, BLUE, and YELLOW pigments. These fake spectral curves look like this:

white_ks

yellow_ks

blue_ks

These approxmiate white, yellow, and blue, but when mixed still create grey:

mixture

To perform more interesting or convincing mixing, better spectral data is needed. Feel free to contribute improved K/S curves if you develop them. Do not contribute curves that may be considered IP.

Running

You can demo the pigment mixing in your browser. For convenience sake, the serve script runs a Python HTTP server in this directory:

$ ./serve
$ # open http://localhost:9119 in your browser

Contributing

Try changing the colors (or adding more) in consts.js. These are interpreted and mixed in the fragment shader in km.js.

Resources

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published