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)
- True Color to GIF conversion (RGB/RGBA input) with quantization and dithering
- 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)
- stream-based output (via callback)
- source-code conforms to the C99 standard
To get started, we suggest that you have a look at our code examples.
examples/cgif_example_video.cis an example that creates a GIF animation.examples/cgif_example.cis an example for a static GIF image.examples/cgif_rgb_example_video.cis an example that creates a GIF animation from RGB data.examples/cgif_rgb_example.cis an example for a static GIF image from RGB data.
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 src/cgif.c, src/cgif_raw.c, and src/cgif_rgb.c. Here the most important types and functions:
// These are the struct types that contain all GIF data and parameters:
typedef CGIF_Config // global configuration parameters of the GIF
typedef CGIF_FrameConfig // local configuration parameters for a frame
typedef CGIF // struct for the full GIF
typedef CGIFrgb_Config // global cofinguration parameters of the RGB GIF
typedef CGIFrgb_FrameConfig // local configuration parameters for an RGB frame
typedef CGIFrgb // struct for the full RGB GIF
// The user needs only these 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
// The user needs only these functions to create a GIF image from RGB data:
CGIFrgb* cgif_rgb_newgif (const CGIFrgb_Config* pConfig);
cgif_result cgif_rgb_addframe (CGIFrgb* pGIF, const CGIFrgb_FrameConfig* pConfig);
cgif_result cgif_rgb_close (CGIFrgb* pGIF);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 attrFlags and genFlags in the configurations CGIF_Config and CGIF_FrameConfig (or their RGB counterparts). 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 (alpha channel)
CGIF_ATTR_NO_LOOP // run GIF animation only one time. numLoops is ignored (no repetitions)
CGIF_GEN_KEEP_IDENT_FRAMES // keep frames that are identical to previous frame (default is to drop them)
CGIF_FRAME_ATTR_USE_LOCAL_TABLE // use a local color table for a frame (not used by default)
CGIF_FRAME_ATTR_HAS_ALPHA // frame contains alpha channel (index set via transIndex field)
CGIF_FRAME_ATTR_HAS_SET_TRANS // transparency setting provided by user (transIndex field)
CGIF_FRAME_ATTR_INTERLACED // encode frame interlaced
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)
// Flags specific to the RGB API:
CGIF_RGB_FRAME_ATTR_INTERLACED // encode frame interlaced
CGIF_RGB_FRAME_ATTR_NO_DITHERING // disable ditheringIf you didn't understand the point of attrFlags and genFlags and the flags, please don't worry. The example files are all you need to get started and the used default settings cover most cases quite well.
An example can be compiled and tested simply by:
$ c99 -o cgif_example -Iinc examples/cgif_example_video.c src/cgif.c src/cgif_raw.c src/cgif_rgb.c
$ ./cgif_example
Alternatively, you can use the Meson build system:
$ meson build
$ cd build
$ ninja
$ ./examples/cgif_example
In the folder tests, we provide several testing routines that you can run via Meson.
$ meson build
$ cd build
$ meson test
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).
Releases of cgif follow the semantic versioning scheme as described here: semver.org
The following additional guarantees are provided:
- Public API of versions 0.x.x are stable.
There is a Visual Studio Code debug configuration with launch targets for the examples. You need to install the C/C++ extension and a LLDB extension (debugger) to debug cgif.
Licensed under the MIT license (permissive).
For more details please see LICENSE