Prompt Design Resources for MidJourney AI

by Shane McGeehan

Blog: Recent Entries

Newest Posts

Joseph Niépce’s Chickens

Most photographers out there are well aware of the grainy black and white image taken by Joseph Nicéphore Niépce, famously known as the first photograph. What is lesser known (and not true at all, completely made up for the sake of humor) is that the image is not simply the side of buildings outside his window as we once thought, but rather a beautiful chicken farm that Niépce had lived on. As we zoom out from the original Heliograph we can see Niépce’s house, chicken coop, and all the wonderful hens of 1827.

Okay, so I am admittedly a huge photography geek. I went to art school for it, and even taught digital photography to college students. I really wish I was still teaching so I could use AI to mess with them, but a blog post will do… I just wanted to have some fun with the new Photoshop Beta and outpainting an old historical photograph. Might have to try some Daguerre, Talbot, or even Bayard next. Lol.

(click to view)

Photoshop Generative Fill

Adobe just released a new Beta update adding an amazingly powerful tool to Photoshop called Generative Fill – and I stayed up way too late playing with it. I was actually shocked at the quality of the Firefly AI, but also the speed and ease of use. You just make a quick rough selection, type words, and in no time you have 3 options to pick from. Want more? Smack the button again, and now you have 6 options. You can even use the Generative Fill tool with other powerful selection tools. Rainy day? No problem. Click Select Sky, Generative Fill, type “blue sky with fluffy white clouds” – done.

I admit how excited I am about this new feature, and also am astonished at how well Firefly does, but I wanted to see how I could use this new tech along with my MidJourney obsession. Nothing can compare to MidJourney’s models, but they also don’t currently have inpainting or outpainting features. We are artists, so why not utilize multiple tools?

I created an illustrative treehouse scene in MidJourney V5.1 (using my new fun “discombobulate” prompt) then brought the image into Photoshop Beta to extend the canvas, adding more forest. I then got carried away and added the moon, a white wolf, mushrooms, then extended the canvas downward and added a red pond with a bridge. I could do this all day. Omg it’s fun… If you have Adobe Creative Cloud, I highly suggest you install the Beta Photoshop (a separate program) and bring your midjourney images in for some more flexibility.

(click the prompt to copy it!)

/imagine prompt: a cabin in the forest:: discombobulate::2 --v 5.1

(click to view)

Discombobulate

With AI imagery, anything you do to a prompt changes it. Moving things around, even the addition of a comma or something. I love, however, when you stumble upon a term that is REALLY full of magic, even if kinda silly and gibberish-ish.

I am a photographer. Many months ago I tested the term “Knolling” and it blew my mind. Knolling Photography is when you lay objects in an organized manner, like a form of product photography, but when you combine something random with Knolling – things get really fun (and funny).

Well, today I stumbled upon a new magical word… Discombobulate! A term I rarely use in real life, typically meaning to disturb or confuse, creates the most complex surrealist images I have ever seen, similar to some gibberish prompting, but with really interesting results when paired with a main idea. I found that you can add a high weight of this term to any idea (just add “:: Discombobulate::2” to any idea) and things get really busy, silly, and nonsensical. It’s like a weird drug induced hallucination with hints of Hieronymus Bosch and Dr Seuss. Seriously though, you all need to try it!

(click the prompt to copy it!)

/imagine prompt: a city street in the winter:: discombobulate::2 --ar 16:9

/imagine prompt: a daguerreotype of a graveyard at night:: discombobulate::2 --ar 16:9

/imagine prompt: a circus of muppets:: discombobulate::2 --ar 16:9

(click to view)

VR Sculpting to AI Imagery

I always seem to be attracted to things that feel like magic (AI is a clear example of this). Five years ago I got my first Virtual Reality headset. I never tried one, but when I saw videos about Google Tilt Brush and Masterpiece Studio and Adobe Medium, I needed to have it. Today I run everything wirelessly, but back then a long wire suspended from my ceiling with a series of retractable pulleys, wrapped around the room, and went into my gaming rig. It wasn’t ideal at all, but it was unlike anything I’ve ever experienced. VR was exactly what I wanted – pure magic. In an instant my boring, empty, dark room could be transformed into a beach at sunset, a multiplayer battleground, or a studio with endless clay.

I got really into virtual sculpting (like staying up until 3am with 6 hours straight in VR). It felt so intuitive, I mean, I get to use my hands to manipulate material in front of me. I can make it bigger, smaller, and walk around it as if the virtual object is actually in my room. I created a lot of stuff for fun, and even was hired to sculpt.

Now that AI is my new creative hobby, I wanted to see how this technology could take in things I’ve sculpted and make something new of them, maybe more detailed than I was able to sculpt. Here are three examples – an octopus, large format camera, and a circuit board. Not only was I shocked by the similarities to my own creations, but it got me thinking how we might soon be able to quickly sketch 3D objects and “paint in” more details via artificial intelligence. The entire game could change… I can’t wait until I can put on a VR headset, sculpt a quick asset I need for a project, then use AI technology and enhance the looks of my creation. Can you /imagine that?

(click the prompt to copy it!)

/imagine prompt: https://s.mj.run/0EV72P50DXo an octopus under water with bubbles and volumetric light --ar 128:95

(click to view)

The Moth

I am working on a big video project using MidJourney imagery, a flicker film about nature and altered states. It is likely going to take a while (dozens of scenes ready to edit) but I wanted to share one work-in-progress because I thought it was a pretty cool example of what you can get out of a single prompt.

AI is simply a tool, and one you can combine with many other tools. This scene was made by masking out the moth in multiple images, sectioning it into wings, body, and background in Photoshop, then brought into After Effects to animate the flicker effect as well as the moth flying. It is not complete yet, but I plan to have multiple scenes to create a narrative (background music is just a temporary place-holder for what is to come, just a quick edit of me on piano and Terrence McKenna speaking).

(click the prompt to copy it!)

/imagine prompt: photograph of a multicolored luna moth flying in a lush forest at night neon lights psychedelic volumetric lighting neon specks of light --ar 16:9 --v 5b

(click to view)

Big Boat

I am an art photographer and I live by the beach. As one would expect, this means I am sitting on literally hundreds of sunsets and ocean shots. Too many. I can’t ever go out without my camera, and so the nature photographs just pile up on my hard drive. Since I began using MidJourney back in the V2 days, I have been wanting to merge the two mediums, but I never really got inspired until now… I loved this empty ocean image, but wanted to add a whimsical twist. After some prompt experiments, a lot of boats, some upscaling, and masking in Photoshop – here is a bit of reality mixed with a bit of fantasy. I hope it makes you feel happy.

(click the prompt to copy it!)

/imagine prompt: photograph of a large wooden ship carrying multicolored wood cabin style houses patchwork sails bright blue background realistic cinematic --ar 3:2 --v 5

(click to view)

Prompter Typeface

Woah, I have been busy, but I’m finished. A new version of Prompter is now live. Go grab a copy HERE for free (to my supporters, thanks for all the donations, your kindness really makes me smile).

Live music is a passion of mine. I have seen Phish more times than I can count, grew up going to music festivals in my formative years, and I just love that moment when a large group of people all feel the same incredible joy as a whole. Since Prompter now has a new “Typeface Design” batching feature, I decided to make a fun announcement image of Prompter at a concert. A few re-rolls and photoshop masks later, and this was the outcome. Amazing how AI can speed up your creative flow.

(click the prompt to copy it!)

/imagine prompt: https://s.mj.run/r5aWDd6CGfI the letter P photograph of concert lights beams lasers crowd smoke volumetric lighting vibrant color --ar 4:5 --v 5b

(click to view)

Art Photography to AI

I am a media editor. For money, sure, but I mean as a person. It is one of my many passions. I have edited probably 2,000 plus videos (commercial, social, interview), a shameful amount of photographs (nature, art, real estate) – and somehow it has never lost its appeal. I always seek things that I find to feel magical: Artificial Intelligence, Virtual Reality, OneWheel! I think there is magic in editing, like, I can turn something into something else. My hand is in it… I love to shoot – but most of the joy I get is in front of my PC working up the final piece.

Okay, romanticized chatter aside, I wanted to test MidJourney V5 by describing a photograph I made. I have done this with all the old versions and many did (what I thought was) quite well, but I picked a weird edited art image from my recent work (on the left) this time to toss the AI a strange curveball. I will say, I was incredibly shocked at how well it handled the prompt, but also the many other beautiful variations that I am not showing here. I picked this one (on the right) because it was similar to my vision, but also a new thing.

By the way, this post is not some comparison of AI to photography, or about skill or anything like that – just a fun example of seeing how two different creative mediums can… create.

(click the prompt to copy it!)

/imagine prompt: photograph of a line of solid black rectangle columns coming out of the ocean at sunset and stretching into the distance dark blue water orange sun magenta light vibrant colors --ar 3:2 --v 5

(click to view)

Cable Management

This is Edgar. Edgar hates cable management… I will (happily) admit that I am the opposite of Edgar. I am lazy at some things, for sure, but a neat freak about others. I might or might not have a folder on my desktop called “SORT” with months of screenshots I know I will never sort. When it comes to the neatness of my workstation, however, for the amount of stuff I have hooked up to my computer, it’s all very well hidden and compact. It’s a sit-stand desk too, so I worked that slack out.

Anyway, I made this image and portrait with MidJourney AI thinking about someone who was the opposite of me. Don’t be like Edgar.

(click the prompt to copy it!)

/imagine prompt: photograph of cluttered hoarder decrepit warehouse, computers monitors printers speakers, wires extension cord knotted tangled::1.1 bearded long hair old man wrapped in wire cord::1.2 ultra-wide angle lens:: vivid color::1.3 --ar 2:1 --v 5

/imagine prompt: photograph of cluttered hoarder decrepit warehouse, computers monitors printers speakers, wires extension cord knotted tangled --ar 2:1 --v 5

(click to view)

Bangin’ Prompts

Is this a joke? Obviously, yes.
Does it kind of make you think about things? I think so.

I’ve been pondering prompting a lot recently, and how sometimes it doesn’t matter if your prompt is short or way too long, has commas or parentheses, contains nonsense, or is even incredibly well thought out with detailed terminology – it almost doesn’t matter, because you WILL get some results that are interesting. AI is just really good at making an intriguing composition.

Folks who don’t like AI think you can bang your head on the keyboard and make “art” and do they have a point? Yea, kind of. You can (literally) smash a bunch of keys, spell things wrong, not actually know what a word you used actually means, and in the end you might get an image that is incredibly beautiful or surreal… but that’s not what we do, well, most of us.

Don’t get me wrong, I love silly gibberish prompting for fun, but the real work in AI is experimenting with structure, terms, and ideas to get something we envision, or explore things we can’t. I can bang my head on a keyboard, and I did, but control and exploration are what I am really after – and that does take a lot of time and skill. There’s no denying that.

(click the prompt to copy it!)

/imagine prompt: nby hjugt76 vcfd m ,jkn vc v --v 5

(click to view)

Blog Search

Blog Categories

Blog Archives

Big Boat

I am an art photographer and I live by the beach. As one would expect, this means I am sitting on...

Cable Management

This is Edgar. Edgar hates cable management... I will (happily) admit that I am the opposite of...

The Film

People think AI is simple and you just type stuff - sure, sometimes. I have been working on this...

Don’t Use GTP

A little warning on this post. Usually I am a nice and helpful guy, but time to time I need to go...

Willow Dog Runs

I have a little corgi puppy that I adore, but she is a bit crazy. We like taking her for hikes in...

Pin It on Pinterest

Share This