A command line tool to create glitchart from jp[e]g's. The script will save the glitched image.jpg as image_glitched.png bby default. The reason to save the result as png is for stability so the result looks the same on different platforms.
It requires zlib and libjpeg. Refer to the original PIL readme for more info. http://pillow.readthedocs.org/en/latest/original-readme.html
It uses elements of this javascript implementation. https://github.com/snorpey/glitch-canvas/
You should install it in a virtualenv by running pip install git+https://github.com/Kareeeeem/jpglitch or by cloning the repo and running
python setup.py install.
python jpglitch.py input.jpg
There are a few optional parameters
| Flag | Description |
|---|---|
-a, --amount |
This determines the hex value that is used to overwrite original values in the image data. Value from 1 to 99 |
-s, --seed |
This determines where in the image data the script starts overwriting data. Value from 1 to 99 |
-i, --iterations |
This determines how many times the script overwrites data. Value from 1-115 |
--jpg |
Normally the script will output to .png for stablity. We're pretty much corrupting files and unstabe jpegs are rendered differently by different platforms/browsers. Png is more stable, but you might want to keep doing stuff to it so use this if you want a jpeg. |
-o, --output |
Specify the output filename. Defaults to originalname_glitched.extension |
When the optional parameters are not given it generates random values in the allowed ranges. This is closer to the original purpose. Glitches are technically not engineered but they just happen.
python jpglitch.py -a 70 -s 4 -i 31 --jpg -o garbled_mess.jpg input.jpg
