From the course: Adobe Firefly Essential Training

Creating your very first text prompt - Firefly Tutorial

From the course: Adobe Firefly Essential Training

Start my 1-month free trial Buy for my team

Creating your very first text prompt

“

- [Instructor] It is time to try out our very first prompt. Here in the text to image screen I'm ready to start typing in my prompt here at the bottom of the screen. Let's try out the following, A dog reading a book. I'm going to hit enter or return to hit the generate button. And now Adobe Firefly is going to return not one but four individual results based on my text prompt. Now, on the left hand side, while I'm waiting, there are a lot of different options available here in this list, and we're going to go ahead and focus on all of these options throughout this course. Now my first results have now returned to me here in my web browser, and I have four very different images of dogs reading a book. I have different breeds of dogs, different locations, different poses. Clearly this dog seems to have a hand holding the book, which is a bit odd. This one is just enjoying life over Christmas apparently. And this one, again, is just enjoying the outside air. So when you specify or not specify something as part of a Firefly prompt, you are actually opening up the possibility for Firefly to fill in the gaps. Where is the dog? What is he doing? What breed? Where is he sitting? What kind of a book is he reading? What are the circumstances, et cetera. Now let's make a few changes to the prompt and also to the aspect ratio. Now, currently here on the left hand side, the aspect ratio is set to landscape. But what if I'm looking for an image of a dog that is a portrait size? Well, simply click here, choose portrait, and nothing happens. What happens is you basically preconfigure the text to image module and then you go ahead and you click generate. And only then will it take into consideration whatever settings you've applied. So this is not a live editing tool. This is an environment where you have the ability to preset all your settings and then you say, go and now give me that specific prompt result. So this is the next iteration. Clearly someone is holding the book for the dog, which I think is fun. This one is super cute. Again, the more Christmas stuff happening, and this looks a little bit similar to the one we had here just on top, but what does stick out is the fact that these are now all different sized aspect ratios. So this is now using the portrait aspect ratio. Let's get a little bit more specific. So let's say a doc reading a book, space, while sitting at a desk by candle light, comma space, while wearing glasses. Let's make it a smart dog, comma space, with a library of books visible in the background. Let's hit enter or return to generate a new set of images. So now I am also specifying what the dog looks like, what the circumstances are. So clearly there is candlelight and I'm also specifying what the background of the image looks like. So this is something that is really fast. It gets results back really quickly here. And now clearly I can see that I have four very smart dogs as a result of this specific prompt. Now also know that you can also use other types of words like mysterious or cozy. And these are things where we've not really defined what that might look like. So if you type in something like it is a mysterious dog, what does that mean? I mean, is he wearing a hood? Like does he have an outfit? What is he doing? What does the word mysterious mean? And that is open for interpretation, and that's part of the fun using Adobe Firefly. I love looking at Firefly, you know, and just call it like happy accidents, just like we used to have with Bob Ross's paintings, basically. And so this is something where you have a lot of creative freedom to decide what it is you want to do and what these images should potentially look like. So the moment you have your four results return to you, you can click any result to see a larger version of it. Use the arrow keys to cycle through these four results or even use the arrow keys on the keyboard by the way. And when you're ready, just hit the escape button or close to go back to your results. Now, every time you hover over an image, you have a few different options available. I think the most important one here is this one, which is the download button and this one here on the left hand side, which allows you to take your image further. So there are a few Adobe Express related options available at the bottom and a few more prompt related options available here. For example, generative fill or using this image as a composition reference, all features that we will be discussing and teaching you throughout this course altogether. Now, in the rare case where you really do not like the results that was returned to you, you can always provide some feedback here for Adobe. So let's say I don't like this image at all, you can click the thumbs down icon and you can basically explain what it is that is wrong with this image. Maybe the quality was poor or there was some violence featured in here, or maybe there's a copyright violation. You never know what gets in there or what might accidentally resemble something that you might be familiar with. Whatever it is, you leave your notes here and then you submit that information to Adobe. And there's one thing I do want to call out here regarding Firefly. Adobe does not train on your prompts and Adobe does not train on any of the images you might upload manually into this system. The only way for Adobe to see what you're doing is if you report the image, provide information, and only then will your prompt settings and your feedback be seen by an Adobe team, which I think is kind of interesting to know. And there you have it. Our very first prompt looking at four very, very smart dogs.

Contents