If such a renderer would be used for something "serious" (a game with simple 3D elements running in the terminal, for example) the display creates some pretty i

If such a renderer would be used for something "serious" (a game with simple 3D elements running in the terminal, for example) the display creates some pretty interesting constraints on the content creation side. In your video, from 0:33 to 0:38 roundabout, the object "dissolves" in some unreadable kind of blob. On its appearance, it feels like the object can be kinda identified, but after a bit of rotation it's gone. When it starts to grow later on, I at least have the illusion of being able to interpret what I see. Now, this is obviously to the low-res nature of the display, and on that front is not much to change (assuming it has to run in a text terminal). Except using braille fake-pixels maybe, but given the situation like it is with this renderer, what needs to be considered when creating 3D models? I'd think, you probably want as often (as in, from every angle) a clear silhouette, and probably one should keep the on-surface details low in number, not too small in size, and properly elevated/engraved. In other words, rendered conventionally with pixels, it would look hideous :)