Spaces:
Running
Calico Cat
I tried some of user ThunderJames' style prompts to make a realistic calico cat photo. It seems to have ignored the 'full body' and 'wide angle' keywords.
Full prompt: "HD, high resolution photograph, shot with Sigma f/ 2 , 28 mm sharp lens, wide angle (subject= calico cat) + (subject detail= accurate full body, detail fur, detail eyes, detail ears, detail legs, detail tail)"
Second attempt with tweaked prompt: "(HD, high resolution photograph, shot with Sigma f/ 2, 28 mm sharp lens, wide angle) + (subject= calico cat) + (subject detail= accurate full calico cat body, detail calico cat fur, detail cat eyes, detail cat ears, detail cat legs, detail cat tail)"
This calico is ready for her close-up. Still no full body though.
With a simpler prompt, the cat's face suffers horribly, but there's the full body view.
"(high resolution wide angle photograph) + (subject = calico cat)"
There is a step b step process that has to betaken notice mistake:
-->> (HD, high resolution photograph, shot with Sigma f/ 2, 28 mm sharp lens, wide angle) +
instead of :
--->> HD, high resolution photograph, shot with Sigma f/2 , 28mm sharp lens,wide angel : (Subject= .......
Faulty input leads to as I said before either Ignorance or miss interpretation
Problem = HUMAN ERROR!!
Im also just learning about lenses so most of the settings are still experimental and just trying out different Parameters and their effect
The AI also does not appreciate eyes
Yeah, the eyes definitely seem to be a problem unless the prompts are quite specific. Even then, they're often asymmetrical.