I've been trying to get ChatGPT to produce photorealistic images, but they always end up looking kind of cartoony, even when I specify 'hyper realistic.' I've seen a lot of impressive, realistic-looking artwork shared here, so I'm curious about what I might be doing wrong or how to get better results.
2 Answers
Using 'hyper realistic' might not be specific enough. Think about adding details that are common in real photos, like specifying 'natural light' or 'candid shot.' This could help guide the generation process more effectively.
I read that if you check out examples of hyper realistic images online, it might help you craft your prompt better. Maybe use descriptors like 'natural light' and 'analog camera' instead of just 'hyper realistic' for more nuanced results.
Yeah, using more specific terms can make a huge difference!
Thanks for the tip! I'll definitely try changing up my wording.