GaryChildress opened this issue on Jul 08, 2023 · 5 posts
RobynsVeil posted Thu, 13 July 2023 at 5:38 PM
Echoing Gary here - coming back to Poser, but for a specific reason. As I do, I've run down the AI rabbithole with some keenness and alacrity, upgraded my GPU to an 8gig card (3070) which now I find a bit underpowered to run Stable Diffusion - at least, the version I want to run, which is 1.5 - and creating persistent characters for this little novella I'm writing. The only thing is, whilst I'm able to get my characters to *generally* obey my prompts:
facing camera, full body shot:1.125, elegant beautiful 30 year old woman, dark (short) bobbed hair, wearing (ivory blouse), wearing dark skirt, (open mouth), (glaring) at the camera, sitting in (outdoor cafe), outdoor cafe, coffee cup,
which gives me this:
...getting the pose right is a bit hit and miss. And this is where Poser comes in. Yes, there is this site which does a great job with poses, but you're stuck with what they have, which may or may not be suitable. So, I'm having a go creating my own poses for my little scenes so my characters are doing precisely what the scene calls for. Tools: Stable Diffusion 2.1 and ControlNet.
Monterey/Mint21.x/Win10 - Blender3.x - PP11.3(cm) - Musescore3.6.2
Wir sind gewohnt, daß die Menschen verhöhnen was sie nicht verstehen
[it is clear that humans have contempt for that which they do not understand]