image-diet is a Django application for removing unnecessary bytes from image files. It optimizes images without changing their look or visual quality ("losslessly").
It works on images in JPEG, GIF and PNG formats and will leave others unchanged. Provides a seemless integration with easy-thumbnails app, but can work with others too.
App was written and is being maintained by Marko Samastur (email@example.com) and is licensed under MIT license.
WARNING: Development of this app has stopped and moved to backwards incompatible image-diet2.
INSTALLED_APPS = ( ... 'image_diet', )
Check which tools are already installed by executing:
python manage.py check_diet_tools
Install those reported missing or disable them as described by command's
requirements.txt lists all tools together
with their home addresses and tips for installation.
If you are using recent version of easy-thumbnails, then you're done. 'image-diet' will automatically squeeze unnecessary bytes every time a thumbnail is created.
If you aren't, then read further.
image-diet is used to remove unnecessary bytes from images. This means
every byte that will not change final display of the image including meta
information stored in EXIF etc. DO NOT use this app if this is not
acceptable or if your image storage is not a local file system.
Primary motivation for its development was seemless optimization of images
easy-thumbnails. PIL is in many ways a great library, but its
output tends to be verbose.
If you are using a recent version of
easy-thumbnails, then you shouldn't
need to do anything more than described in
Installation. It is important
to disable tools that are not available since app for efficiency reasons
doesn't check during runtime.
You may use
manage.py check_diet_tools action any time to check current
status of external utilities. Action also provides copy&paste ready list of
configuration options for disabling those that could not be found.
You may still be able to use
image-diet even if you are not using
easy-thumbnails. Installation procedure is the same, but you will need
to trigger shrinking from your code (or let me know which public app you are
using so I can add support for it).
To do this import:
from image_diet import squeeze
path_to_image is an absolute
path to image you want to optimize. Function returns
None if there was a
problem or path to squeezed image if it was successful.
Returned path is currently always the same as the one provided, but this may change in the future (when GIF to PNG8 transformation gets added).
If you installed
image-diet after you already processed some images, then
you can shrink them with
manage.py diet_images command. Just pass paths
to directories you want to scan for images as command's argumentand it will
process all images that can be found in those directories or their
DIET_DEBUG = True
This will keep uncompressed versions of images on disk with
.diet. Defaults to
DIET_<TOOLNAME> = False
Disable use of
TOOLNAME. Name has to be written in uppercase so
DIET_JPEGOPTIM = False will disable jpegoptim. Defaults to
- add extreme compressions (change to b&w, reduce color depth, change GIF to PNG8)
- add support for storage other than local file system
- stop depending on tools that processed image will actually be smaller
- add integrations for other image handling Django apps
- app doesn't check if files exist so some operations could lead to data loss (if image folders contain files with .diet or .orig extension)