Skip to main content
Participant
December 27, 2023
Question

Neat, but does't seem to understand context very well

  • December 27, 2023
  • 1 reply
  • 264 views

I've been playing around with Adobe Firefly today, annd while I really like it and can see the potential, it has also proven to be quite frustrating, as it doesn't seem to understand context very well. I signed up for the Firefly monthly plan. 100 credits didn't seem like a lot, but I figured it would be enough for a trail month. I uploaded a photo that I wanted to edit and started playing around. Certain things were easy, like changing eyecolor. Other things the AI just didn't seem to understand, such as adding facial hair in a particular color. I tried adjusting the settings to align more to the prompt vs the source image, and it didn't really seem to matter. I've now burned through about 75% of my monthly credit allotment, on just the one image. I've tried other photo editors, such as Canva and pixlr, which seem to understand prompts better. However, Canva specifically warns against using on faces, which is specifically what I was looking for. I'm hoping Firefly will improve over time.

This topic has been closed for replies.

1 reply

Kevin Stohlmeyer
Community Expert
Community Expert
December 28, 2023

Hi @Joel344821979o76 the X factor in all of this is writing effective prompts to control the expected outcome of the AI. The good news is Gen FIll credits aren't actually enforced until after Jan 1 last I saw.

This page may help guide you with writing more descriptive prompts:

https://helpx.adobe.com/firefly/using/tips-and-tricks.html