Corgi is a performance-focused fractal rendering engine capable of ultra-deep zooms and highly flexible styling.
Corgi uses a combination of several precision-extending techniques to render views of the Mandelbrot set at nearly infinite zoom levels. So far, the deepest rendered image had over
Corgi uses optimized algorithms and efficient hardware utilization to render images as fast as possible on your hardware, often achieving real-time interactive rendering. Features include:
- Parallelized rendering in GPU compute shaders
- Fine-grained caching to avoid re-rendering when unnecessary
- Immediate UI feedback combined with asynchronous re-rendering
Corgi includes several coloring algorithms and a layer-based compositing system to allow an incredible amout of variation even for the same fractal locations.
Corgi is still in alpha, so there are many more features I still plan to add. Until version 1.0, the exact algorithm and save file format may have breaking changes.
Above about
Internal coloring algorithms are unstable at high zoom levels, and internal distance estimation is not implemented yet.
Currently, binary releases are not being created. However, you can install Corgi by cloning the repository and running:
cargo install --path .This will compile a binary named corgi.
Basic CLI options can be viewed with --help:
corgi --helpIf the application fails to load, it likely encountered an issue during GPU initialization. Open up an issue, and I will see what I can do to support your machine. Currently, I have tested on dedicated GPUs on the Vulkan backend.
To get more information about what is happening, you can set the CORGI_LOG_LEVEL environment variable:
# run with the highest level of detail
CORGI_LOG_LEVEL=TRACE cargo run --releaseCorgi is distributed under the Apache License (Version 2.0). Some code is adapted from other sources under the MIT license. The corresponding portions of the code have been annotated in the source.
See LICENSE for the Apache License, and here for the MIT license.