Feature Requests

The current beta UI for speech generation is not viable for projects longer than a few minutes. Even a 5 minute project that takes advantage of the range of features currently available could easily an unmanageable number of files, and turn what should be a straightforward process into a tedious task of wading through bad edits & duplicates created in error in search of the 6 or so that represent your completed work.Instead, implement a text-block work flow where the system accepts the entire project text, parses it into text blocks (sized as willed) which the user then works with in a hybridization of the captioning conventions (merging/splitting text blocks, editing or adding text, etc.) and the speech generation features (voice, speed, pitch, tone/emotion, etc.). Additionally, allow the users to preview partial blocks or bridges between blocks, and to generate individual, selected, or all blocks.I believe this model of UI maximizes the opportunity to apply current and future edi

Under Image 2, the ability to Edit > Generative Fill only allows for download; there is no clear UI method to return to previous results, which means the prior images get re-rendered when using the browser back button.   It would make more sense to open the Edit environment in a new tab or as a frame to allow closing the environment without reloading the previous page.

une fois le images générées par le texte, il serait interessant de pouvoir l'ajouter dans la bibliothèque ( pour ceux qui dispose d'adobe creative cloud)

pretty much just like some other ai generators that do not have filters which censor ridiculous words like gun

It would be great that when I return to the tab in my browser that had the prompt I was working on, stay there, or reload automatically from when I last visited the tab. Currently the tab will reset back to the default page and I lose what prompt I was working on. 

Hello there.Hereis the requirement that I have and currently cant do using firefly. I am trying to generate few pictures of models wearing jewelleries. Now I want them to wear jewelleries, pictures of which that I will upload. I have tried to use the reference picture, howver it does not work at all.

It would be create if you could define the image type to be like a clipart/vector image and set it to say 4 or 6 colors with clear boundaries between the colors.As an example logos.This would make it possible to print the results on a capable multicolor 3D FDM printer.Kind regards

"download of images to library not working when trying to type nam it disappears 

We cant select image ratio before generating the first picture

As I was using the tool and I starting marking some of the results with the "Star Icon" as a favorite, I noticed that there was no way for me to go back to that list of favorites to look at the bookmarked/saved assets.Maybe an icon could be added to the top next to idea generator so that users can easily find their bookmarked/favorite asset list 

Potreste implementare la funzionalità di creare dei video dall' Ai

Having AI understand how to create Cubic Environment Maps, Reflection Maps and Textures for 3D application would be great. Textures seem to have moderate sucess already, but it would be good to have a way to define scale of the texture. I think "Seamless Tile" already works, but if not, that is helpful too.

It will be great if we can submit an image from our own collection to be used in the graphic generation process.

It would be interesting to animate a PNG or Jpeg logo with FireFly. Imagine a one colored or Duo colored logo that would "melt", "bounce", "crystalize", or "explode into frame" by uploading this PNG or JPEG and the generative AI would handle the keyframes allowing the user to export a 1 ~ 2 second animation of their logo. Nice idea 😉

Sehr geehrtes Adobe Team, ich würde mir sehr wünschen wenn man mit einem Text, sich mehrere Bilder erstellen lassen kann, dann die Bilder auswählen kann und sie zu einem kurzen Video erstellen lassen kann. Die KI soll dabei zwischen den Bilder einen sanften übergang erstellen das es so aussieht als würde das bild sich nur verändern und nicht mehr aussieht als wären es mehrer Bilder an einander gereiht.   Mfg   [Removed by Mod]

It would be helpful to have the option to generate one image instead of four, especially in the beginning while you are still finding your way. This way you can get directions on your idea quicker and it will put less strain on the servers.

I would like to be able to see my previous creations and the prompts i used to make them. thank you

Generative fill could be even more useful if it could interpret prompts of "same thing but..." to make alternations to what's already in the image instead of replacing it.  For example, "same face, but eyes looking left and mouth open," or "same armor but more damage."

And for the roblox character I choose, I say make it fit the background and improve the lighting, and I do the same for various creative comparable backgrounds that are not random and are precisely what I want. This is for the generative AI and creating comparable backdrops to the other thumbnails, as well as utilizing the character I chose and improved and another model to create lighting or something similar.

make it an app

How about a Microsoft Teams integration? I'd love to use Adobe Firefly within Teams.

We should be able to go from text to image into the generative, fill and back out into the text to image again and continue to do this to keep on refining. If this is already a feature, it's not clear on how to use this to go back to continue working with the other filters. It seems like a one-way street.. if somebody has direction on how to accomplish this please inform me I am really enjoying using this and I feel it just needs to be tweaked a little bit more to get things more user-friendly I'm sure the trillion dollar corporation can afford such workers to get this done. 

It would be great if we could go back from the generative fill editing back to the previous window/space (e.g. text to image) without losing our images, settings and prompts.

When I use the generative fill module, I would like to see the difference between the original image and the edited one. Would you please add a compare button to see the original image?   Please also add a back button to reverse the changes Thanks [Removed by Mod]

Is there a way to upload a photo and have it converted to a new graphic.  I want to use the photo attached with the cat and turn it into some type of fighting leprachaun similar to the sample image attached.