Lossy compression faster than fourier transform based methods at similar ratios.

[Gabe Kontros]

I'm not sure how this discussion page works. I've made a lossy compression algorithm that runs mostly off extremely cheap and basic arithmetic once per pixel. (one subtraction, one addition, and two Look Up Table accesses). I also use chroma subsampling. I want it to be free and open source and was wondering if anyone working on AV1/2 or AVIF would be interested. Functions mainly off of the discrete derivative. There's also still a lot of room to grow. I want it to be free for anyone anywhere to use or improve.