Text to image

Transcript

We are now going to look at text to image to start. Click the text to image link here. As you can see, there are several examples of what other users have generated using Adobe Firefly. You can hover over them to look at the prompt they use as this could be a really useful way to guide you in writing your prompt. However, in this instance we are going to simply create a prompt from scratch. It’s important when you are writing prompts that you are as descriptive as possible. Words that have an abstract meaning to them don’t tend to work. So for example, something like the word efficient, you would need to describe what you mean by the word efficient because Firefly will struggle with that. In short. Visualize your idea with text. Once you type in your prompt press, generate and wait for Firefly to do its magic. You will be presented with four options. Once we have our first results, we can now start refining them by using the menu on the right hand side. The first option is the model version that we have available at the moment. Currently, Firefly version two is selected, which means that the images will be much more realistic and much more crafted. Firefly version one creates more stylized images, so it’s up to you to choose which one you want. Depending on the look you desire. Next, you can change the size of the generated image from the dropdown menu. So for example, I want this to be widescreen so I can then use it in my PowerPoint presentations. You can also choose to decide where you want the result to be more photographic, which is what I want in this instance, or something a little bit more creative. Visual intensity determines how close to being realistic or not. The output is moving on. If we press generate again, Firefly will change the size to match what we’ve selected and increase in visual intensity. If you like one of the images generated and want Firefly to generate results based on that click edit and select show similar, which means that it will retain that one generated image and generate variations based on that. Next, we could match the style of our generated images to a reference image. You could upload an image from your library so the image generated matches your brand style. For example. You can also use images in the reference library. For example, you can create a hand-drawn image, a photographic image that is a bit more dramatic, or for example, something a little bit more creative like watercolors. Let’s test one of these more creative outputs and see what Firefly generates. Once you choose a style reference, it appears in the bottom left corner. Press generate for some results. The results produced will match the style reference. We’ve now got an image that is a little bit more hand-drawn and beautiful that we can use in our presentations if we need to. However, if I wanted to use an image based on my brand guidelines, I could simply upload my reference. Before you do so, make sure you remove the previous style. Also, make sure that the image that you are using is one you have the rights for. Once you’ve done that press generate. Now Firefly will generate images using the prompts based on the style reference, and this will allow you to generate images that are on brand. As you can see, the results have a dark look to them, and if we compare them to the original image, they look very similar. Other refinements you could make to the results generated include applying some effects. You can choose to make them more like a painting, hyper realistic, simpler, flatter, etc… These effects will change the results generated. Another aspect you can change is the color and tone. Maybe I want this to have vibrant colors or be a little bit more monochromatic, golden and so on. Maybe I want this to have a specific lighting design. For example, you could choose to have dramatic lighting. Lastly, you can choose the composition as well, directing how the photograph is set up. So for example, you might want more surface detail to be a bit more present, a little bit more close up, etc… With our choices made, we can now generate new images. Once you’re happy with the image, you can either download it, copy it to generate a link to share with other users, or copy the image itself so you can place it in any other applications such as PowerPoint.