Image Creation issues - over processed

[Petra_Lentas]

Anyone else getting unusable image outputs from ChatGPT image generation lately?

I’ve been using image gen (DALL-E via ChatGPT) for design work — print-ready dog illustrations for apparel. For months, it was great.

Now?

• Every image is cropped (ears, tails, paws missing)
• Faces are vacant and weirdly distorted
• Worst of all — every output is sharpened aggressively, turning soft sketch styles into crunchy, unusable messes

Even images that look great during the render preview come out over-processed when finalized. No setting to control it. No way to stop it.

This started happening about a week or 2 ago and has gotten progressively worse to the point that i have been unable to generate a successful image for at least 4 days. Is anyone else seeing this? Or did I miss a model update that quietly broke everything?

[polepole]

Hi welcome to the community!

Can you share your exact prompt and its result?

[Daller]

The generator is using a nonsensical detail and sharpens “improver”.
Is it this you talk about?

Collection of GPT-4o-images prompting tips, issues and bugs

The new system is not a continuation of DALL·E 3 but a reboot. It produces a completely different style that resembles generators like Flux more than the old DALL·E 3 model. It also seems that the developers consider the images too flat and have added something like a structure enhancer in the final phase. However, this often worsens the results. The images may become sharper and seemingly more detailed, but at the same time even darker than they already were, and above all, the changes lead to…