Hi everyone, I'm pretty new to using AI image generators for work. I've been trying DALL-E 3 through ChatGPT to make some simple graphics for our internal project updates and social posts.
Every single time I ask it to put any text on an image—like a banner that says "Q4 Launch" or even just a simple sign with "Open" on it—the letters come out all jumbled and mixed up. It's like alphabet soup! For example, I just tried “A modern logo with the text ‘SwiftCRM’ below it” and got something that looked like “SwiffttCRmM” with random symbols.
Is this a known issue with DALL-E 3? I see so many amazing images here, so I figure I must be doing something wrong or there's a trick I'm missing. I'm using pretty straightforward prompts. Any guidance would be super helpful 😅
Yeah, it's a huge known thing with DALL-E 3 and most image generators right now. They're terrible with text. It's not just you.
I've found you can sometimes get clearer results by describing the text placement in extreme detail, like "a sign with clearly printed black letters that spell OPEN, with perfect spacing". But even then it's a gamble. It's faster to just add the text myself in Canva or something after the image is generated.
Anyone know if Midjourney or other tools are actually any better at this?