https://github.com/mncaudill/3bitdither

Dithering that runs completely client-side
https://github.com/mncaudill/3bitdither

Last synced: 3 days ago
JSON representation

Dithering that runs completely client-side

• Host: GitHub
• URL: https://github.com/mncaudill/3bitdither
• Owner: mncaudill
• License: mit
• Created: 2012-04-01T06:30:56.000Z (over 12 years ago)
• Default Branch: master
• Last Pushed: 2013-08-05T15:57:06.000Z (over 11 years ago)
• Last Synced: 2024-08-02T05:11:37.034Z (3 months ago)
• Homepage: http://mncaudill.github.com/3bitdither
• Size: 53.7 KB
• Stars: 23
• Watchers: 0
• Forks: 12
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        

3 Bit Dither

==============

Implemented by Nolan Caudill

This demo currently uses two different error-diffusion dithering algorithms: Atkinson's and the Floyd-Steinberg algorithm.

Error-diffusion means that the algorithm goes pixel by pixel, rounds the individual R, G, and B channels to either 0x00 or 0xff, and then distributes those differences (which the algorithm calls the "quantization error") in differing amounts to pixels further down the line. It being a 3-bit dithering means that your red, blue, and green channels are represented by a single bit (off or on), giving you a total of 8 colors.


The demo also does halftone dithering and Bayer dithering (just 2 bit) which don't use error-diffusion, which for these two makes the patterns evident that error-diffusion usually lessens.


This runs completely client-side, using the FileReader and canvas APIs. If you have a somewhat modern browser, this should work. Also, you can right-click and save the result of the dithering.


The demo lives here.