Bug Reports

Hi,I`ve just started to use Firefly, but everytime i ask for a picture, it shows me mountains and fjords.Ive tried to exclude theese words, but it still shows me theese pictures when i try to generate something. Can you help me? Thanks. 

Lately the platform is not giving me high quality results or correspondence to the prompts

 I get this message every time i try to use text to image, any suggestions will be helpful. Thanks

When I upload an image of a full figure. The generated image is missing the top and the bottom of the render. I've been going back and using auto generation with the crop tool, which is an extra step that I would rather not do. Anyway, just wanted to let you know.

I always land on the home page of Firefly, and it does not give the option to pick what size you want to create until you hit generate, then the next screen opens to the interface where you can pick sizes, filters, etc. It uses 1 credit that I didn't want to use, but had to in order to get to the next screen. 

iOS: 17.4.1Browser: Safari Wenn ich auf like oder dislike bei einem Generierten Bild Klicke fürt das dazu das ich nur noch mit diesem Bild interagieren kann. Es wird auch automatisch beim Klicken auf eines der anderen Bilder zu dem Bild gewechselt wo man den like, dislike gegeben hat. Dies istfür mich eindeutig ein Feler und persönliche schrenkt es mich ein.

So, right before 3 minutes or so I was trying to generate a gothic realistic photo of a pathway, the prompt that I wrote is: Overgrown cobblestone path to abandoned village reclaimed by nature. Crumbling ivy-choked stone cottages/houses. Looming Gothic church consumed by vines. Eerie blue-green bioluminescent glow from mushrooms/foxfire. Still, silent air with sense of dread. Drifting mist casting ominous shadows. Unseen lingering presences. Unfortunately Adobe Firefly didn't generate it and it even took my token, I tried changing the prompt a little but that didn't work either. I think that's a bug. its like giving you money to buy something but you're getting nothing but still giving your money. 

Structure ref of my own creation of a cartoon person keeps triggering user violation!

When I ask your AI to make a taino warrior in traditional taino clothing it gives me a man with a sombrero. and big guitar along with traditional bright mexican colors. I don't how an AI can be racist but the Taino people of the caribbean are indeed NOT mexican mariachis. Copilot gave me some great, accurate, and respectful options. Please train the AI better or say that it can't generate the image rather than doing something so audacious.

For a project, I needed to represent caucasian people. And your program is ONLY able to generate Black and Asian people. I understand this is trying to be inclusive but it defeats the point when you aren't including all the races. You end up being exclusive. You can do better. I tried multiple prompts and its unable to generate any white person. Come now.

I have noticed for a while now that Extend Image and 3D to Image have been in the "Try these out" section, where only available, public production features that can be used are shown - but these two features have been and still are disabled and in the exploration state. Normally, they would be in the section dedicated to not yet available, WIP features, at the bottom of the page - which confuses me if it's just a bug and these features are actually publicly available and useable, or if it's, as it says, still in exploration and it's just a UI variable or flag incorrectly set making it appear in the wrong section. 

Olá, ultimamente tenho utilizado a ferramenta Remover para retirar objetos da imagem e ela gera sempre no final uma mancha, parecendo um arco-íris. Esse bug já vem acontecendo tem um tempo e me impede de realizar as atividades com o Firefly

The landscape images produce results wherein the fourth image is impossible to view fully or download in the Firefox browser. Tried refreshing the page, it did not help. Please, I'd prefer if you can revert to the original layout with all images visible. 

the image was not downloaded and the credits also over only one image edit please increase the credits for accounts iam poor thats why i cont buy the abedo 

Todos os arquivos no preenchimento generativo estão completando de forma incorreta, deixando uma parte branca ou com algo nada haver. O recurso de completar imagens está muito ruim! Se continuar assim irei cancelar a assinatura.

Nothing I do can make cabin lights at night look dark. Cabin windows at night always have lights.

Antes no era necesario hacer tantos intentos para lograr una imagen óptima, ahora comete muchos errores.Ver captura de Video:https://www.loom.com/share/3dc34e1428084127a1fc3e34c5d969ab?sid=c0b329e5-20f6-48d4-a2fb-a11b4d958b67

Newest version of Firefly doesn't allow for mouse/trackball/touchpad pointer-scrolling - as in no scroll bars appear to move view of the list of settings.Using Google Chrome (ver 123.0.6312.86) on Windows 10 Pro PC.  Never had a problem before, now it's an issue.  It makes it harder to change settings for the AI image output.  I do NOT have a touch-screen.

O preenchimento generativo, mesmo que configurado "Preservar conteúdo > totalmente novo" e "Força da orientação > totalmente novo" está preservando cerca de 90% da imagem original.  Ao "esticar" a imagem no computador e fazer o upload, esse  "esticamento" é uma barra toda branca, que vou deixar transparente no firefly, para preenchimento. Antigamente, quando se clicava em "gerar" sem adicionar nenhum prompt, o firefly preenchia com perfeição todo esse espaço. Agora, ele mantém 90% de fundo branco e só adiciona uma transição entre a imagem a ser mantida e o fundo branco, que deveria desaparecer.

Trying to extend the background, but it's not working.

I tried to add generative fill to add a dinosaur to my landscape photo.  I tried it numerous times and it created weird dinosaur images with no head and tails on both ends, created images that were nothing like the known dinosaurs, etc.  A total fail.  So disappointed and a very basic, simple prompt.

Why does every person have to have a stoopid smile? Even when I type "Sad", they still smile.

Extra layer and vertical lines are showing while replacing the entire background. Both the normal picture and the generative fill picture are attached.

When I select the background, it does fine. But when I click generate on the ai tool, it generates images that change the subject too. Most frequently turns it into a black blob instead of just changing the background.

I uploaded an image of my own invention. Meaning I invented the product in the image. It's mine as is the photo. Why am I getting this error: Your reference image violates Firefly user guidelines and was removed. It's my product called, The Bee Raft. They save bees from drowning in backyard water sources.