I like the idea of "squint to see." I also jotted down an idea to use an image that's made of some type of material in order to have like a double meaning, but I still haven't ended up using it for some reason. I think these types of 2-images-in-one renderings are super duper neat, though.
Also: your knowledge of Midjourney has to be in like top one percent of users. Have you interacted with other MJ geeks along those lines?
Yeah the "squint to see" illusions were all the rage for a while back in mid-2023. People aren't going as crazy about them now, but they're definitely still neat to play around with.
I haven't talked to any people directly, but I do stumble upon Facebook groups or Reddit threads discussing different approaches, and since I'm interested in this space, I usually pay attention.
Remember stereograms? These were images you'd see in places like malls during the 90s, and you had to sort of blur your eyes to make two sets of dots come together. I wonder if gen AI is sophisticated enough to make one of these.
I think the prompting part has actually only gotten more straightforward now: Midjourney's pretty good at understanding natural language, so you can pretty much just describe your desired scene to get close to what you need.
But yes, it also has all of these additional levels of control that let you edit and tweak images beyond text prompting.
Fantastic Tips!
Glad you found them useful! Let me know how it goes if/when you give them a whirl.
I need to reboot my midjourney membership to try but I'm socking them away :)
I like the idea of "squint to see." I also jotted down an idea to use an image that's made of some type of material in order to have like a double meaning, but I still haven't ended up using it for some reason. I think these types of 2-images-in-one renderings are super duper neat, though.
Also: your knowledge of Midjourney has to be in like top one percent of users. Have you interacted with other MJ geeks along those lines?
Yeah the "squint to see" illusions were all the rage for a while back in mid-2023. People aren't going as crazy about them now, but they're definitely still neat to play around with.
I haven't talked to any people directly, but I do stumble upon Facebook groups or Reddit threads discussing different approaches, and since I'm interested in this space, I usually pay attention.
Remember stereograms? These were images you'd see in places like malls during the 90s, and you had to sort of blur your eyes to make two sets of dots come together. I wonder if gen AI is sophisticated enough to make one of these.
Here's at least one anecdotal example from last April: https://www.facebook.com/groups/officialmidjourney/posts/556552473302988/?paipv=0&eav=AfbhWredmiNE8B9LolpFK3Nfb8Q2w6e96SM-if5vCcU2kuIFhF17gh7p1c8HFVklil0&_rdr=
Not bad, although of course it would be more impressive to see an actual shape.
Fun. Thanks.
You got it! Feel free to share any observations if you end up trying these out!
Glad you enjoyed it!
I think the prompting part has actually only gotten more straightforward now: Midjourney's pretty good at understanding natural language, so you can pretty much just describe your desired scene to get close to what you need.
But yes, it also has all of these additional levels of control that let you edit and tweak images beyond text prompting.