Click the flag and simply watch. This is a 9 frame long gif I converted to be made out of lines. Probably going to encode (or have someone else encode) the Pokemon intro or something before releasing the source code, especially with how much of a mess it is. It's also written in JS which makes it take forever for each frame -- Nearly 100 seconds per frame here. (Still not much slower than using circles with @MartinBraendli's settings (80 seconds per frame), to put how slow JS makes this into context). Update: I've done it. Took 24 hours of encoding. https://scratch.mit.edu/projects/563021622/ I've now encoded it at a 480x360 resolution for both the output and the gif, which slowed it down nearly 4x, which I really didn't like (but I'll not be at my computer for a bit so it won't be too bad) -- 240x180 was just enough that the algorithms could sometimes find tiny details which was good enough for this first display of it.
480x360 encoding that I accidentally forgot to record the stats for and will most likely re-encode to do so: https://turbowarp.org/560659948 Pokemon the Series: Ruby and Sapphire for the scene this tiny gif is from. @Finlay_Cool for creating this project that partially inspired me (and helped me find @MartinBraendli's project because I just can't find it otherwise -- I personally like this one's look more though) https://scratch.mit.edu/projects/34367372/ @MartinBraendli's source code for a basis of how I'd create this out of lines. https://scratch.mit.edu/projects/72880262/ This list size is actually smaller than what you would get from their project! This is because while they generate videos with 512 circles, each one requiring 1 color (giving 4 values for circle x, y, size, and color), this uses 341 lines with one color (giving 6 values for line x1, y1, x2, y2, size, and color). 512*4=2048, and 341*6=2046. The settings for this gif was: 240x180 image processing 341 lines per frame 50 random lines (half of the time circles were generated instead to allow for tiny details to be found) 150 random mutations. The errors for this gif: Maximum error: 528347.13, being for the first frame Average error: 451122.11 Minimum error: 406642.39, being for the eighth frame Compare this to an attempt using circles with 's settings (240x180 processing, 512 circles, 70 random circles, 40 mutations) which are: Maximum error: 583028.99, being for the first frame Average error: 487463.48 Minimum error: 438954.96 being for the eighth frame
