Dither

Inspired by: http://www.tannerhelland.com/4660/dithering-eleven-algorithms-source-code/

• no dithering:
• onebit black and white floydsteinberg dithering
• CGA 16-color mode, sierra 3 dithering:

License: MIT

CLI help:

bash dither --help

usage

bash dither INPUT [OUTPUT] [-v] [--dither] [--color] [--depth]

example:

```bash

dither "bunny.ong", output to green_bunny.png, using the atkinson dithering algoirthm, and using two bits of depth for shades of GREEN, emitting debug info to stderr.

dither bunny.png green_bunny.png --dither=atkinson --color=GREEN --depth=2 -v

dither "bunny.png", output to bunny_crayon.jpg, using the sierra 3 dithering algorithm, and loading it's color palette from crayon.plt

dither bunny.png bunny_crayon.jpg --dither=sierra3 --color="crayon.plt" ```

args

INPUT

the path to an input image. known good image formats are .png and .jpg.

OUTPUT

the path to write the output. this will create or truncate the file at the existing location, if necessary. the image type will be inferred from the filename. currently, only .png and .jpg are supported.

-c, --color

The color mode to use. Default is "bw" (black and white). Color mode to use.

"color"

(all colors with specified bit depth)

"bw"

black and white mode (grayscale in higher birt depths)

$COLOR

single-color mode. options are

• BLUE
• GREEN
• CYAN
• RED
• MAGENTA
• BROWN
• LIGHT_GRAY
• GRAY
• LIGHT_BLUE
• LIGHT_GREEN
• LIGHTCYAN "LIGHTRED"
• LIGHT_MAGENTA
• YELLOW
• WHITE

$FILENAME

load a palette from file. palettes are specified as a list of two or more newline-separated 24-bit hexidecimals, with optional 0x prefix. see crayon.plt and cga.plt for examples. // WHITE 0xffffff // BLACK 0x000000 // RED 0xff0000 // GREEN 0x00ff00 // BLUE 0x0000ff

-d, --dither

The type of ditherer to use. Available options are

• "floyd", "steinberg", "floydsteinberg" (default, floyd-steinberg dithering)
• "atkinson",
• "stucki",
• "burkes",
• "jarvis", "judice", ninke" Jarvis-Judice-Ninke dithering
• "sierra", "sierra3" Sierra dithering

-v, --verbose

Verbose debug output to stderr

--depth

Default 1. Bit depth should be an unsigned integer between 1 and 7. The number of bits to compress each channel to. This option only works with the options --color=$COLOR, --color=bw, --color=color