Skip to main content
Participant
March 25, 2023
Released

P: Transparent background or layers

  • March 25, 2023
  • 114 replies
  • 12300 views

Is there a way to generate an image with a transparent background. Altrnantively, will there be a way to produce an image as file with layers so we can easily  remove the background? 

Pinned Reply By droopydog500

This can be done in the Firefly AI Assistant:

    droopy

114 replies

Participant
June 1, 2023

it seems possible to remove the background easily but i cant then download that image without the background which i might want for something like a game sprite

 

Merging your thread here so we can track the demand for certain features. You can upvote the original post to let us know you want this feature.

Participant
May 27, 2023

please enable download of the transparent png of the subject after removing background automatically .

 

 

 

generated new background usually destroys part of subject like fingers or similar. Having subject cutout would help later to overlap witg generated image nad have neat and clear subject inage in the final render.

 

Merging your thread here so we can track the demand for certain features. You can upvote the original post to let us know you want this feature.

Participant
May 27, 2023

It would be amazing to have the chance to make images with a transparent background and download them as PNG

 

Merging your thread here so we can track the demand for certain features. You can upvote the original post to let us know you want this feature.

 

Participant
May 15, 2023

an option to generate an object as a png with a transparent background

 

Merging your thread here so we can track the demand for certain features.

Participant
May 10, 2023

If is possibile...can be interesting export file .psd with layers?

For me will be very interesting

 

Merging your thread here so we can track the demand for certain features.

davidpreli
Participant
April 17, 2023

Hello there, 

 

It would be interesting to beable to program with text blendmodes either in intial image generation or while using a previously generated image as a reference. 

 

I created the the attached image in photoshop by taking two text-to-image creations and overlaying one ontop of the other with the linear light blend mode and a transparenscy of 42% with some masking around the mantis. 

Cheers,
David

 

Merging your thread here so we can track the demand for certain features.

Participant
April 15, 2023

Would be great if was possible to create multiple paralel prompts as layers for the composition. For instance an image isolated from the total composition that you can move around and scale overlapped with multiple similar instances. This way we'll have way more creation possibilities and control of the final imagery. I'm imagining something similar to Adobe Photoshop Layers Menu UI, but where each layer have a prompt/text commaand for AI and some feature to select and remove the "background" of each layer, similar to the Magic Wand. I hope it can be easily understood, I tried my best haha 🙂

 

Merging your thread here so we can track the demand for certain features.

Participant
April 13, 2023

If particular objects could be generated on layers, this could be even more transformative. This could use the same layering framework that Adobe Express also uses to make everything simple. Saving these layers out separately would be very cool also, or saving the file as a PSD like smart object.

 

Merging your thread here so we can track the demand for certain features.

Participant
April 14, 2023

Love the idea. I wish to be able to lock certain apsects of results (background/foreground/character/etc) and if these things were on layers, it seems like it would enable this.

Tracy.H
Community Manager
Community Manager
April 12, 2023

@Humberto292256112vme Merging your thread here so we can track the demand for certain features.

 

Participant
April 12, 2023

When selecting graphic or art, the out out should have a minimum of three layers.  Background, subject, and foreground.  AI is amazing, but will always need to tweaking from Us humans.  This would give us ability to fine tune the output and potentially move to proper position some items.