AI art generators have been in the news a lot this year, whether for their amazing advances or questionable uses. OpenAI’s Dall-E 2 is one of the major names in this space. It is now open to the public and developers, and will soon be built into Microsoft software and the Bing search engine.
Stock art service Shutterstock is also set to integrate the tool and pay artists it copies to return them and (perhaps) avoid some ethical issues. After all, Shutterstock’s art was used in part to train the Dall-E AI.
But how exactly do you work with Dall-E? Is it really as simple as writing a description – called a challenge – and getting an image back? To be honest, yes. However, if you want to get closer to the perfect result, you need to be aware of much more. Let’s talk about it.
Register and pay
The first thing you need to do is create an account at labs.openai.com(Opens in a new window). Use a Google or Microsoft account or create a login with an email address and a strong password. There is no multi-factor authentication option.
Dall-E is not completely free. The service runs on “credits(Opens in a new window).” You get 50 free credits when you sign up, then 15 free credits per month, but they won’t roll over. Paid credits roll over from month to month for up to 12 months; get 115 credits for $15.
One credit allows you to create one generation of AI art (you get four new images for a typical generation). This might start with a challenge, but also credit is used to create a variation of already created art. You can use a lot of credits trying to hack the correct image generated by the AI.
The perfect challenge
Once you have registered, you will be confronted with a form for your challenge. If you click on Surprise me button, you’ll see more random prompts inserted into the text box – they don’t count towards your credits until you click generate. You can also upload your own image and use Dall-E to edit it and add new AI-generated content or create completely new variations of the original.
But the main thing is the challenge. It’s likely that most people will come across. We did our own experiments to see how random challenges might work with Dall-E and its competition like Midjourney(Opens in a new window) create art. The initial challenge always does something…interesting. But it’s rarely perfect. Things are always a little mushy. A bit strange. A little off. The art will show up much better if you perfect the challenge, which is limited to 400 characters (emojis(Opens in a new window) work(Opens in a new window)).
Design Challenges(Opens in a new window) have both content (what you want to see) a modifiers (how it should look like). For example: “A robot draws a picture on an easel” is the content, but “over-the-shoulder view, colorful, oil paint, in the style of Van Gogh” are all modifiers.
Not feeling very creative? There are people out there to help you create the right challenge to get the right art. They go by job titles like Prompt Designers or Prompt Engineers, and their expertise doesn’t come free. But if you’re going to do it a lot, getting the challenge right up front can save you money because you won’t use up as many credits.
In addition to videos like the one above, there are resources like The Dall-E 2 Prompt Book(Opens in a new window) from dall-era gal-era. Worth a read. It explains many “hacks” to consider in your challenge, such as using terms like “close-up” and possible camera angles, types of lighting, a list of eras to emulate, such as “20 years’ or even mentioning the specific type of camera or smartphone lens that ‘takes’ the AI image.
Using emotional words to get a more positive or negative image is big, as is using words to set a mood / aesthetic(Opens in a new window). The challenges of getting the most out of AI-generated “photos” and “illustrations” are a world unto themselves. The possibilities are endless. Your only limit is 400 characters.
The only thing you can’t do with Dall-E is create pictures of real people. Well, you shouldn’t anyway. It has filters and restrictions in place to make sure, but with billions of images to study, who knows when Dall-E might release an image that looks like a celebrity. This means you can reference movies and shows to approximate their look and feel.
Can you request a style that mimics a living artist? Of course, and if Dall-E had seen this artist’s work, he could probably provide an imitation. A style can’t be copyrighted, but of course it’s a crappy thing to do if you want to use it commercially when you could have worked with the artist.
Any image you generate in Dall-E – or any image you upload to Dall-E (make sure you own the copyright) – can instantly get variation. Uploaded images must be cropped to a square image with a 1:1 aspect ratio.
You can also create variations of the variation. Just click the menu with three dots on the images and select Variations to do more.
The variations really cannot be controlled. It’s simply Dall-E looking at the content of the image and creating something roughly the same, with the same styles, similar layout, etc.
Recommended by our editors
Let’s say you made an image with Dall-E and you like it. Mostly. But something is not quite right here. Choose Editand remove the part you don’t like with the eraser tool, rewrite part of the prompt to address that part (say a face that needs a more specific description, or a whole new background to replace the original). Dall-E will create a variation based on the new specifications in the challenge. Your prompt must still describe the entire desired image, not just the deleted area.
This eraser option also works with images you can upload, so you can remove the head or the entire background. Write a prompt for what you want to display in the deleted section. For example, erase part of the sky and use the “inverted plane to crop dust” prompt to see it appear overhead. Working within the limits of the generated image in Dall-E is called “inpainting”.
Delete part of the uploaded file, create a challenge and generate new AI art at that location. (Credit: PCMag)
Make the image bigger
So what is “painting”? In general, output from Dall-E is limited to downloading a 1024 x 1024 pixel image. Another thing you can do under Edit is to create generational frameworks. Click the Add Generation Frame icon, which looks like a frame with a plus sign in the top left (or press the letter F), and you’ll get a floating frame that you can place anywhere outside the image’s perimeter.
Frame generation for Dall-E image extension. (Credit: PCMag)
Click Generate again and the image will expand into that frame as if the AI had just continued to draw/photograph into the expanded space with the same challenge. Keep doing this until the image is as big as you want. Of course, each new frame generated consumes credit.
When you’re ready to own it, click the download button (down arrow) to get a .PNG file of your new masterpiece. After that, if you really want to increase the image resolution, you can try a site like bigjpg.com(Opens in a new window) or ARC Face Restorer(Opens in a new window) (which also corrects faces) or commercial software such as the $99.99 Gigapixel AI(Opens in a new window).
Remember, you can always return to the My Collection page on the Dall-E website to view all the images you’ve ever created, including variations. So if you had an idea to improve an older image that you never downloaded, you can go back to it.
If you’ve created something really awesome, you can click Share > Publish. This will place the image on a public page on openai.com where you will get a special link(Opens in a new window) share with others. You will see the image plus the prompt used to create it. If you don’t want to go public, try it Save > Favorites put in a collection that you can view later.
Become an artist
Are you ready to experience the art of artificial intelligence? All you have to do is register and write. Try it out on Dall-E with free credits and you’ll have a taste for all the AI-generated art we’re likely to see in the next few decades.
Variation for upload, expanded with new images. (Generated on Dall-E)
Like What are you reading?
Log in tips and tricks newsletter with expert advice to get the most out of your technology.
#DallE #art #generator #create #amazing #images #text