I played with DALL-E from OpenAI when it first came out. There are some nice examples on their site but in my experience it’s pretty uneven.
Sure, sometimes you’ll get an astonishing painting that looks really cool…but most of the time, the face will be borked, someone will have three hands, or the software will hilariously misinterpret your intent.
On the other hand, Midjourney seems to consistently produce phenomenal images.
For example, “Tony stark on his laptop getting ready to buy something and smiling mischievously” yielded this:
Not bad at all. Not sure I’d interpret that as “mischievous” but emotions are notoriously tricky to represent in still art.
Half the fun is playing with various prompts, such as
- “Cyberpunk runic letter for a dystopian maximalist propaganda. High detailed, 8k, monochrome, black and white, minimalism, brutalist architecture”
- “Pixar style german shepherd dog nuzzling up to a smiling teenage girl”
- “superhero taking selfie running from the police, vivid and intense, fisheye lens, 8k”
- “draw a diagram of many different IT components (some cutting edge cyberpunk, some very old steampunk) interacting with each other in complex ways and often failing”
Which yields the following:
Cyberpunk runic letter for a dystopian maximalist propaganda. High detailed, 8k, monochrome, black and white, minimalism, brutalist architecture
Pixar style german shepherd dog nuzzling up to a smiling teenage girl
superhero taking selfie running from the police, vivid and intense, fisheye lens, 8k
draw a diagram of many different IT components (some cutting edge cyberpunk, some very old steampunk) interacting with each other in complex ways and often failing
It’s quite entertaining and once you get going on an imagine, there’s all kinds of ways to refine it and tweak it, or select from the initial images and refine.
The app is Discord-based, and they’ve discontinued their free trial. However, it’s pretty cheap. The lowest tier is $10/month. It’s cheaper if you buy for a year, but…you’re dealing with a company that didn’t exist a year ago so signing up for 12 months might give you pause. That tier will give you about 3h20m of compute time which is more enough to make thousands of images.
As I said, it’s Discord-based, so after signing up and picking a plan, you log into their Discord and start issuing /imagine commands. T
Hmm, let’s try this…
“laptop in nature, iceland, iceland, nature, realistic, details, mountains, green grass, blue lake, gray sky, 4k, high resolution, general geographic, fog, , elegant, hyper realistic, super detailed, photography, cinematic, intense, cinematic composition, , beautifully color graded, Unreal Engine, Cinematic, Color Grading, Editorial Photography, Photography, Depth of Field, Tilt Blur, White Balance, 32k, Super – Resolution, Megapixel”
Wow.
Related Posts:
I also enjoy using LEBs! I have an empire of little guys for VPNs, larger guys for apps, and big guys for databases and storage. When I'm not in front a screen I'm into German Shepherd dogs, quality knives, target shooting, theology, tabletop roleplaying games, and forest hiking.
I enjoy writing technical articles here on LowEndBox to help people get more out of their systems. You can find me on LowEndTalk @raindog308
- One Week From Tomorrow…THE WORLD WILL LOSE THEIR MINDS!Lines Are Already Forming! - November 21, 2024
- Crunchbits Discontinuing Popular Annual Plans – The Community Mourns! - November 20, 2024
- RackNerd’s Black Friday 2024: Bigger, Better, and Now in Dublin! - November 19, 2024
Pretty much my experience:
“Sure, sometimes you’ll get an astonishing painting that looks really cool…but most of the time, the face will be borked, someone will have three hands, or the software will hilariously misinterpret your intent.”
Have not taken the time to use midjourney, looks I got to give it whirl.
The second face looks like a real person.