OK I had the distinct impression that there were now tools that could “rasterize” images to characters (similar to #AAlib and #libcaca), but with more aggressive usage of #Unicode characters to achieve better results. I have tried #img2txt from libcaca with the #utf8 option but it's still not satisfactory compared to what I remember, even for grayscale, so … which one was it?
Also #TIL that this is called #semigraphics or #pseudographics
I have found not one but TWO #Python things called img2unicode that promise to do something similar to what I want to do. One of them is little more than exercise, and has questionable results and performance. The other seems to have been curated more and can even be installed by pip, but is also overengineered for what I care about (which is only to leverage the Unicode Blocks set, whose characters has a relatively well-defined “graphics value”).
Do I have to roll my own?
It's quite obvious I'm not the only one pursuing these kind of interests, even if not always with the same aim:
Attached: 2 images I made a neat terminal-friendly cross hatching effect with unicode
Oblomov