P: Text and symbol generation in images

[Angelic_child5CF3]

I am trying to generate a logo with a desired text. It is not accepting it.

[Angelic_child5CF3]

Text and symbol generation in images is not supported yet in the Text to Image module. The quality of our models will improve with time, and generated content will better respect the prompt. If you are trying to create just text, please try the Text Effects module. (Merging all similar threads here so we can track the number of reports for this)

[yasmin_9717]

Eu solicito para o firefly desenhar um logo escrito Street Worship, mas sempre erra a escrita do "Worship"

 

 

[droopydog500]

Hello @yasmin_9717,

Thank you for your message. I am sorry you are having this problem. 

While it has gotten better, Firefly (and most other AI image generators) is generally not good at putting exact text from a prompt into an image. Sometimes it is perfect, but usually, it does not get it right.  The best solution is to use Adobe Express or Photoshop to add text in.  The more complicated the angle or perspective of the text, the more likely you will need Photoshop to get what you want.

Please see this article on why generating text in AI image generators is hard:

If AI image generators are so smart, why do they struggle to write and count?

Thanks,
    droopy

[Serge_I&D7185]

no escribe la palabra, la cambia

[droopydog500]

Hello @Serge_I&D7185,

Thank you for your message. I am sorry you are having this problem. 

Is the problem that your text is being converted to English? If so, what is happening is that when it detects a prompt in a language other than English, it translates the whole prompt to English before it generates. This unfortunately causes any text you are trying to insert to be translated as well.  The best work around is to write your prompt in English except for the text you are trying to insert.

If you are having a different problem, please explain in more detail. Please show examples, if possible.

Thanks,
    droopy

[Kartika Rawat]

Hi Serge_I&D7185!

Sorry about that. Are you trying to add text to the generated images, and the text is not generating correctly? If so, we know that there are artifacts in some scenarios. However, text and symbol generation in images isn't supported yet in the Text to Image feature. You can provide feedback or report images with distorted or unclear text to let us know. In addition to this, you can try Adobe Express to add text to generated images.

Let us know if you have any questions.

Thanks,
KR

[drahomir_5381]

Prečo prakladá aj mená do angličtiny

Chcem vytvoriť srdiečko s menom " A L E N K E " a stále mi to tam dáva Alice

[droopydog500]

Hello @drahomir_5381,

Thank you for your message. I am sorry you are having this problem. 

What is happening is that when it detects a prompt in a language other than English, it translates the whole prompt to English before it generates. This unfortunately causes any text you are trying to insert to be translated as well.  The best work around is to write your prompt in English except for the text you are trying to insert.

My best,
    droopy

[JoAnnSis]

Its a shame. Im not a fan of gates,  but his app does text pretty well. I thought id give adobe a shot. But adobe has a long way to go!!

[youngdubstudios]

how come i cant get firefly to spell correctly? 

[droopydog500]

@youngdubstudios, 

Please see this article on why generating text in AI image generators is hard:

If AI image generators are so smart, why do they struggle to write and count?

    droopy

[kashish_jain]

I want to generate text - Ooze but its incorrectly generating Ouse.

[Oh.N8]

@kashish_jain 

 Ya... this can be a bit of a problem. I find pumping the slider to maxium helps considerably.

Cheers,
O. Nate

[Kartika Rawat]

Hi kashish_jain!

Sorry about this. Currently, text generation in images isn't supported yet in the Text to Image feature. You can provide feedback or report images with distorted or unclear text to let us know. You can generate images without text and use Express to add text.

I hope this helps.

Thanks,
KR

[hugo_3535]

Pourquoi l'application fait toujours des fautes d'orthographe ?? Merci

[droopydog500]

Hello @hugo_3535,

Thank you for your message. I am sorry you are having this problem. 

While it has gotten better, Firefly (and most other AI image generators) is generally not good at putting exact text from a prompt into an image. Sometimes it is perfect, but usually, it does not get it right.  The best solution is to use Adobe Express or Photoshop to add text in.  The more complicated the angle or perspective of the text, the more likely you will need Photoshop to get what you want.

Please see this article on why generating text in AI image generators is hard:

If AI image generators are so smart, why do they struggle to write and count?

Thanks,
    droopy

[Charles32925702e4ms]

When I ask Firefly to generate an image that includes text, the image it generates has the text misspelled.

See attachment for an example.

[Abraham_MotionGraphics]

I asked FIrefly to put the words "Congrats Mo" into an image and it failed miserably.

[jane-e]

@Abraham_MotionGraphics and @Charles32925702e4ms 

AI generators including Adobe Firefly are not good with generating text. See the article posted by @droopydog500 on September 25 earlier in this thread for the reasons.

In the meanwhile, use Adobe Express for text.

Jane

[Kevin Stohlmeyer]

@Abraham_MotionGraphics honestly those results are impressive for an AI to generate. Getting it 90% right makes it possible to edit in PS or other graphics editors to finish the job.

[Monika Gause]

I do not think that you can correct the text in an image editor, but using effects you might get close rebuilding it.

[Kevin Stohlmeyer]

@Monika Gause I didnt say correct the text. You can edit in Photoshop and recreate the missing letter. A lot less effort/time than struggling with a limited outcome via AI.

[Monika Gause]

---

@Kevin Stohlmeyer  schrieb:

@Monika Gause I didnt say correct the text. You can edit in Photoshop and recreate the missing letter. A lot less effort/time than struggling with a limited outcome via AI.

---

Yes, @Kevin Stohlmeyer. that's what I got. But still I do not think that will be simple. The left example from the screenshot can be recreated as a live text effect in Illustrator. That one on the right is a bit harder, but I think you can get close.

[Kevin Stohlmeyer]

All the elements are there - the reason this would work is there already is another letter in the missing "G" space.

Combining the C and other elements to create a G and some cleanup - did this in 10 minutes rough.

[Jacob241703391h0s]

this app cant spelll   why am i paying for a logo maker that cant spell     im gonna post everywhere on social media about this 

strait scam

[tzurlis]

trayed to do a photo using the prompet :

a homeless photographer begging for mony in the street holding a cardboard sighn saying "i will shoot for food scraps"

1. didnt get the sighn right

2. all the homlesses were clean, happy, standing up - like a disny movie. ever seen a real homless person?....

3. cmera was not there or hidden

4. generating time was very long

5. other engens dune a much better job. 

that is pritty shamfull for an adobe size company.