CGIF, a GIF encoder written in C
A fast and lightweight GIF encoder that can create GIF animations and images. Summary of the main features:
- user-defined global or local color-palette with up to 256 colors (limit of the GIF format)
- size-optimizations for GIF animations:
- option to set a pixel to transparent if it has identical color in the previous frame (transparency optimization)
- do encoding just for the rectangular area that differs from the previous frame (width/height optimization)
- fast: a GIF with 256 colors and 1024x1024 pixels can be created in below 50 ms even on a minimalistic system
- MIT license (permissive)
- different options for GIF animations: static image, N repetitions, infinite repetitions
- additional source-code for verifying the encoder after making changes
- user-defined delay time from one frame to the next (can be set independently for each frame)
- source-code conforms to the C99 standard
To get started, we suggest that you have a look at our code examples.
cgif_example_video.c is an example that creates a GIF animation.
cgif_example.c is an example for a static GIF image.
To get an overview of the API, we recommend having a look at our wiki (https://github.com/dloebl/cgif/wiki/General-API) where types and functions are described. The corresponding implementations can be found in
cgif.c. Here the most important types and functions:
// These are the four struct types that contain all GIF data and parameters: typedef CGIF_Config // global cofinguration parameters of the GIF typedef CGIF_FrameConfig // local configuration parameters for a frame typedef CGIF // struct for the full GIF typedef CGIF_Frame // struct for a single frame // The user needs only these three functions to create a GIF image: CGIF* cgif_newgif (CGIF_Config* pConfig); // creates a new GIF int cgif_addframe (CGIF* pGIF, CGIF_FrameConfig* pConfig); // adds a frame to an existing GIF int cgif_close (CGIF* pGIF); // close the created file and free memory
With our encoder you can create animated or static GIFs, you can or cannot use certain optimizations, and so on. You can switch between all these different options easily using the two attributes
genFlags in the configurations
CGIF_FrameConfig. These attributes are of type
uint32_t and bundle yes/no-options with a bit-wise logic. So far only a few of the 32 bits are used leaving space to include further functionalities ensuring backward compatibility. We provide the following flag settings which can be combined by bit-wise or-operations:
CGIF_ATTR_IS_ANIMATED // make an animated GIF (default is non-animated GIF) CGIF_ATTR_NO_GLOBAL_TABLE // disable global color table (global color table is default) CGIF_ATTR_HAS_TRANSPARENCY // first entry in color table contains transparency CGIF_FRAME_ATTR_USE_LOCAL_TABLE // use a local color table for a frame (not used by default) CGIF_FRAME_GEN_USE_TRANSPARENCY // use transparency optimization (size optimization) CGIF_FRAME_GEN_USE_DIFF_WINDOW // do encoding just for the sub-window that changed (size optimization)
If you didn't understand the point of
genFlags and the flags, please don't worry. The example files
cgif_example_video.c are all you need to get started and the used default settings for
genFlags cover most cases quite well.
Compiling the example
An example can be compiled and tested simply by:
$ c99 -o cgif_example cgif_example_video.c cgif.c $ ./cgif_example
Validating the encoder
In the folder
tests, we provide several testing routines that you can be run via the script
performtests.sh. To perform the tests you need to install the programs ImageMagick, gifsicle and tcc (tiny c compiler).
With the provided tests you can validate that the encoder still generates correct GIF files after making changes on the encoder itself.
The GIF format employs the Lempel-Ziv-Welch (LZW) algorithm for image compression. If you are interested in details of the GIF format, please have a look at the official GIF documentation (https://www.w3.org/Graphics/GIF/spec-gif89a.txt).
Licensed under the MIT license (permissive).
For more details please see