Feature Requests

Located in Serbia, Europe.Seems server is not working properly 

Maybe a method to incorporate 3D into Firefly: Stereographic: text to image method1:Set 1:1, hi-res, rgb, and generate the image according to the prompt.2:Cache the results under left_eye.(L)3:Move the camera/lens 2.5” to the right.4:Generate the image using sameseed.5:Cache the results under right_eye.(R)6:Do a comparison of left_eye and right_eye7:Correct the anomalies using a healing algorithm.8:Generate version options for:VR/AR: (L)(R) 1:2 high-res3D viewer: (L)(R) 1:2Cross-eye: (R)(space)(L) 1:3Red/green: isolate the r-channels and the g-channels 1:19: Dsplay selected version(s)

A tool that allows you to specify the direction you would like for 3D Objects to flow in the text.

A combination of Firefly and the Unreal Engine 5 would be perfect. Individual open worlds, houses, interiors, trees, shrubs, etc. would be possible.

You can add feature for how many image user want to generate (like 1,2,3,4).And also you can show us thumb image only in first view for data saving (who aren't using WiFi, it can be pleasure for them like me) and then when we click for any exact image then load the full image. Otherwise you can show us small size preview image.Thank you so much for your helpful tool and I think you will consider my suggestions.

Be able to export the font as a font file to be used in other text programs like word

alpha channels for the objects in the image, allowing the artist to rearrange and work on each one individually. Of course only when it's possible, like figure and backgound, the subject etc.

HiI would like to suggest the "Recolor vectors" would make a palette available for us.It could be 5-20 matching colors with the RGB/CMYK color codes readySometimes a color combination just works, and then it would be great to have the above information.

Would be really great if you had the option to change the color on particular letters.For example for "FLIORG" with Rainbow broken glass, It would be nice if I had the option to make F red, L Orange, I yellow, etc.

I would be great to be able to have a layer option in the Firefly text effects to have more control.Layer order help choose which prompt goes on top or bottom and enable to stack up different effect on to of each other to have a better control of the mix

Hello hope all is well! I am sure you are all working very hard to create this now. I am a musician and it's hard as an independent contract to create high quality music videos that are engaging. It's very expensive. would love to have an AI that can listen to my audio the song and create an actually moving video 2-3 minutes in length. wouldn't need to be anything fancy but some form of art that flows with the beat and makes sense with the lyrics. thank you so much! best regards,Robair  

It would be very useful to have to possibility to create own style and recall it when needed, like flagging a particular set of own generated images that are similar.Then you can upload your "style 1 preset" and create new consequential images: this would be fundamental in helping users create comics and/or manga! Eman  

Maybe an idea to have a text pop up, when you hover over the different styles, with an extra explanation. So its easier to understand precisely what every style does to your prompt. Often i find myself clicking on random things without really understanding what will change.  Or maybe, instead of a text pop up, it shows a gif when hovering over it, just like in photoshop, showing quickly in a global way what the style means.

As a 3D Printer I think having the option to convert your 2D ideas to 3D printable files would be an amazing idea.  

I suggest we swap the order of Thumbs down/Thumbs up buttons in the "Rate this result" tab. I almost hit Thumbs down for an image I liked because I'm used to the regular order I find on Facebook or LinkedIn, YouTube, where the first button (the one on the left) is the Like button (so Thumbs up), and the second button (the one on the right) represents a second option. In Firefly they are swapped and create a bit of confusion for the first few seconds. It's counterintuitive. It would greatly improve user experience if the buttons would follow the regular order: Like (left), Dislike (right).

Would be nice to be able to re-roll just one letter when the text effect is good for all the others. Like in the above, I wish I could just regenerate the V without touching the others.

You have a brilliant way to design stylised fonts, but not a way to make fonts more personalised. Give us the ability to type something in a fornt of our choosing then use prompts to modify that specific font in terms of attributes and appearance,

The effect would be so much better if each of the letters was a unique creation. In words and phrases with multiple repeating letters, the effect of style repetition decreases the value. I tried the text "Bianca's Bees" - the two Bs being a copy of each other, and the two Es together made the repetiveness very obvious and non-attractive. 

The same as the text effect generation but for fonts styles.

Dear Adobefirst of all this is a logo designers dream come true .. wish I would have that in the 90swhen I was designer for several demoscene groups ... really really well done Adobe !!what a great tool it is ... ( using just the text option for now ) I need definetly access to all my fontstext options like in photoshop to change trackingdynamic canvas sizes ( so loose items are not cut away )different resolutionspositioning of individual text / letterswarp effects ( for fonts / letters ) extrude !!! with depth / perspective positioning / light position 🙂  yes yes I know ... ehm rotation maybe too 🙂  ..Allow stronger text effect fits ( loose should be even more loose ... a slider would be cool that I can slide all the way to the right until its not recognisable anymore. layers !! place individual letters on their own layer Overall, I want to say that if AI is used correctly, it can be magical. Firefly is a perfect example of this. Thank you

Would love to get specific with which colors are generated.

It's great to have some basic controls for lighting such as golden hour and studio lighting. But even deeper controls would be helpful, like "lighting from left", "Lighting from right". Most concept artists use some combination of photo elements, 3d and hand painting in Photoshop to arrive at their final concept. The problem with photo elements is they often have the lighting from the wrong direction, and so when you add those elements to your painting they don't match. Would be great to use firefly to edit the lighting in those photo elements to whatever you need. 

When creating materials for 3d applications, you frequently need to make tileable textures from photos. Would be super helpful if you could take an AI generated image (or an uploaded texture map once we're able to use uploaded images) and say "make tileable" and it would make the image tileable. More info on the subject in this video: http://www.neilblevins.com/art_lessons/ai_tiling_texture/ai_tiling_texture.htm

If Adobe Firefly could do this, it would be revolutionary and amazing: generating an image but from two camera angles and rendering it as SBS or top-bottom so it can be viewed in magnificent stereoscopic 3D.Currently, using LeiaPix converter, a fake 3D (depth-map based) image can be created, but it will never compete with a true geometric 3D.See some Adobe Firefly imagery in fake 3D here:Adobe Firefly AI in 3D VR 

Hi there, hope you're doing great!One idea as a suggestion could be the possibility to add custom color since now only we have the option to use a pallet predefined.This could be a good implementation since if someone wants to create something and use their branding color is not possible. I love the tool, I'll keep testing it.Best regards,