didder is an extensive, fast, and accurate command-line image dithering tool. It is designed to work well for both power users as well as pipeline scripting.
It is backed by my [dithering library](https://github.com/makeworld-the-better-one/dither), and is unique in its correctness and variety of dithering algorithms. No online or offline tool I know of provides as many options, while being correct (linearizing the image).
- Yours? You can provide your own error diffusion matrix in JSON format
More methods of dithering are being worked on, such as Riemersma, Yuliluoma, and blue noise. If you'd like to help out with development of those methods, or request a new one, please make an issue in my [dither](https://github.com/makeworld-the-better-one/dither) library repo, not this one.
Download a binary from the [releases](https://github.com/makeworld-the-better-one/didder/releases) page. On Unix-based systems you will have to make the file executable with `chmod +x <filename>`. You can rename the file to just `didder` for easy access, and move it to `/usr/local/bin/`.
On Windows you will have to open a terminal or cmd window in the same directory as the EXE file, or [add it to your PATH](https://stackoverflow.com/a/41895179).
Make sure to click "Watch" in the top right, then "Custom" > "Releases" to get notified about new releases!
### From source
**Requirements**
- Go 1.14 or later
- GNU Make
Please note the Makefile does not intend to support Windows, and so there may be issues.
The best place to learn about how to use didder is the manual. Run `man didder`, or look at the [MANPAGE.md](./MANPAGE.md) file. If you only read about one flag, read about `--strength`. It's especially important for `bayer` dithering of color images.
You can also run `didder` to see the global options and commands. Each command represents a different dithering algorithm, or set of algorithms. You can see the options for each command with `didder help cmd` or `didder cmd --help`.
This command dithers `input.jpg` to just use black and white (implicitly converting to grayscale first), using a 16x16 Bayer matrix. The result is written to `test.png`.
As another example, here's the command used for the image at the top of the README:
If you'd like the replicate this yourself, the input image is available [here](https://upload.wikimedia.org/wikipedia/commons/7/71/Michelangelo%27s_David_-_63_grijswaarden.png).
## What method should I use?
Generally, using Floyd-Steinberg serpentine dithering will produce results with the fewest artifacts. The command would be:
```shell
didder [palette and I/O options] edm --serpentine FloydSteinberg
```
Playing with the strength of the matrix might also be useful. The example above is at full strength, but sometimes that's too noisy. The command for 80% strength looks like this:
```shell
didder --strength 80% [palette and I/O options] edm --serpentine FloydSteinberg
```
The main reason for using any other dithering algorithm would be
- **Aesthetics** - dithering can be a cool image effect, and different methods will produce different and stronger artifacts
- **Speed** - error diffusion dithering is sequential and therefore single-threaded. But ordered dithering, like using `Bayer`, will use all available CPUs, which is much faster.
If you want to see examples of the different dithering algorithms, you can look at [this directory](https://github.com/makeworld-the-better-one/dither/tree/master/images/output). Or try them out yourself!
## License
This project is licensed under the GPL v3.0. See the [LICENSE](./LICENSE) file for details.
