I know I’m a judge, and inherently will not/cannot submit an entry, but here’s how I’d convert an image to ascii. (Even though the contest didn’t say anything about ascii, I really like the idea)
I’d throw a few trained convolutional neural networks at my frames, checking for edges, at angles that can be represented in ASCII. (90 or 270, 180 or 0, and what ever angles the slash and backslash symbols are. I’d also have a few checking for a filled square vs an empty one, etc.)
I’d chuck that output into text file(s).
I would then initialize some form of viewing window, and display each frame at the correct speed.
As for the 8bit graphics, I would most likely have a program that I spent 30 seconds on (to fulfill the suggested 8bit graphics thing) that would convert an image to an 8bit image.