In computational terms, a low resolution version of an image is almost by definition ‘simpler’, with fewer colours and details intact, but it seems like it would be much harder to do a convincing 1:1 replication of it in a painting compared to recreating a ‘clean’ HD version.
Or am I way off the mark? 😆 I’m not a painter, obviously. Seems like getting all of those weird JPEG artefacts right would be something of a novel skill for a traditional painter (or even a digital painter, for that matter).
I mean… pixel art. For sure it is hard in a completely different way than photorealism. Less to work with (for good and bad), and a new set of rules (for pixel placement). I’m sure there’s multiple valid techniques (digital first, rough planning, individual pixel canvases/swatches or some other collage etc… not to mention cross-stitching or various building toys if you count that).
I don’t see the appeal in re-creating artifacts, but I’m sure there are people who can make a convincing approximation (particularly if they know any of the technical reasoning for JPEGs).