00:00 🎨 Midjourney's Describe Tool version 6 transforms images into text prompts, aiding creative processes. 01:49 🔄 New Describe Tool optimized for version 6 generates more complete conversational prompts, offering a creative springboard for various projects. 03:45 📸 Describe Tool excels at identifying iconic locations and famous people, enhancing photography or image prompts. 04:59 🎨 Analyze artworks with Describe Tool to receive prompts with style terms and artists, aiding in understanding and replicating artistic styles. 06:28 🌍 Use Describe to build fantastical worlds by feeding it images from different media and using prompts to develop landscapes, creatures, and cultures. 07:30 🕰 Utilize Describe to glean specific historical details from photographs or paintings, ensuring accuracy in historically themed artworks. 08:02 📝 Let Describe aid in crafting captivating captions for art, allowing the artwork to speak for itself on social media and websites.
everyone has done the --cref tutorial (always traditional hooman types), hope u can do a deep dive into strange looking characters, im not having good results :(
I’ve never asked what prompt anyone has used because I have a pencil and paper. It’s called creativity and process, it’s how you make art that means something. You’re just generating an averaged image of a copy of a copy, you didn’t make any fundamental choices in the composition, contrast, rhythm, etc. it’s a Getty image
Bully for you! You can draw! Technique isn't art. There are plenty of people who can draw who HAVE NO IDEA WHAT TO DRAW. The fun of AI art is in the intellectual creativity. What could we draw -- if we had the technique. Don't be mixing up technique with creativity. Right now the tools are basic, but I can't make choices about composition, contrast, rhythm, etc with a pencil. I can't draw! It's not a Getty image (I'm a professional photographer) You have a camera on your phone, but I bet you can't make photos like I can. It's not the tool! It's the creativity behind the eyeballs.
In alpha, they've kind of broken the offered prompts down into bite size pieces. Instead of generating 4 seperate ones, it generates 4 initial descriptive sentences, then a few artists, if any apply, and then the detail stye bits, so you can pick and choose which of them to combine into your own prompts. EDIT: Imagine all only works in fast mode or turbo.
For now /describe is only in discord. If you have access to the alpha website, there is another sort of describe tool. Drag an image into your prompt and open it. Click the i icon. You will get parts of a prompt you can select.