Prompt Design Resources for MidJourney AI

by Shane McGeehan

Blog: Recent Entries

Newest Posts

Photography to Video, in RunwayML

A good artist experiments and explores, so I try to use all kinds of AI to learn how they function. 😎 I have been toying around with Runway ML lately, which is an artificial intelligence model that generates video clips from text prompts – however – you can input your own images to start with. Since I am a photographer, I wanted to see if I could use this new AI to create a short video from my still images in a very short deadline. Here is the result.

I will admit, it’s not perfect, but it is quite interesting, and shows the direction we are headed with these new powerful creative tools. I am excited to see where Runway as well as others like Pika take this new technology.

(click to view)

Be Careful Out There

Here is a very different kind of blog post. A warning, or something like that.

We need to talk about something really weird that’s happening on facebook. Recently there has been a huge spike in random “pages” posting AI generated images of a person and art with stupid captions like “my wife made this wooden owl with his own hand, appreciate it” or “my son made this and no one likes it, please like” and then, because most people are fucking stupid, the image gets literally thousands of likes and shares, along with boomer comments like “amazing talent” and stuff (sorry, not sorry, because most of the commenters are boomers). Here are a few I collected in the last 48 hours.

Now, it seems harmless, but why would “Life Is Beautiful” or “Garden Flowers” or “Love Dogs” take the time to post awful AI images just for some likes? It doesn’t add up to me. Maybe I am overthinking things, but I feel like this could be either a simple way to gather a list of names and profiles of very gullible people to later scam out of money or passwords – or even darker, what if something like this was a big bait-and-switch plan, and when the US election season heats up “Love Dogs” becomes “Vote Freedom” or something and starts spreading political misinformation or even worst, false political AI images. Again, I don’t know, but it’s freakin’ weird and nothing can be done about it because Facebook is a runaway monster. I hope I am just paranoid, but you never know.

Anyway, be careful out there. Don’t accept everything as truth, and always look at images very closely to make sure they are not fake. Question everything, because as I see it, this is only the beginning.

(click to view)

Evil Shane Styles

Evil Shane I made a custom “Evil Shane” style, and I encourage you all to make some Evil Shane’s and share them if you are up for it, prompt below… but, before you think this is some genius science, it’s not. It’s stupid. I love messing around with new tool, plus it’s fun, and I love fun.

(1) I used the new /TUNE feature with a URL link of a headshot of myself with a simple prompt “portrait” after it. (2) I selected a few really creepy Shane images to fine tune my style. (3) I then ran a bunch of PERMUTATION prompts to find the best looks with variable stylize and image weight. (4) I used the /PREFER OPTION SET command to make –EvilShane replace a prompt with my URLs headshot,style code, stylize, and image weight. (5) Lastly, I pushed through a bunch of random prompts adding –EvilShane to them to get really weird results! (keep in mind the –EvilShane thing is local, for me only, so you’ll need to copy the entire prompt below)

Want to make an Evil Shane? Just type in any prompt idea and add to it:

(click the prompt to copy it!)

/imagine prompt: YOUR PROMPT HERE https://s.mj.run/ZAVpESe9BOw https://s.mj.run/rATstUmIv6w --style buLsr5r9E9IERM-2U4NwPvHtUVQG --stylize 850 --iw 0.8

(click to view)

Fine Tune Your MidJourney Styles

My last video about the Style Tuner was a quick intro, but now we dig deeper and really fine tune those looks. Here is a longer tutorial that goes more in depth, so you can be a pro in no time. Learn about the new Prompter and a few useful features, how to best use the Style Tuner for your work, and follow along with some random and funny prompts.

HERE is a link to the Style Tuner I used so you can play around, and below are style codes for the three example styles in the video. Feel free to copy and use in your own prompts while you watch!

(click the prompt to copy it!)

--style 1GRrA8D6rhfnB2c2nsN1E7euH (bright, washed out, vintage)

--style Hkrz0ovIoeA5q4U0cbR (dark, underexposed, creepy)

--style yVW0VP8Wuj (colorful, pop, surreal)

(click to view)

How to TUNE your own STYLE

Hayyyy… A new MidJourney feature was just released that allows you to fine tune your own styles. Sure, it might seem complicated at first, but watch this little 5 minute tutorial of how I use TUNE and you’ll be a master in no time.

You can use TUNE for a single prompt idea, select more or less pairs to fine tune, or even combine the feature with –stylize to make the style stand out more. In this video I do what I enjoy, using “looks” but you can also be generic and enter “a city street” then select only 1 or 2 of the styles you like best, and run some prompts. Want to try more? Just unselect those and pick another 1 or two, and run more prompts with your new style code! – I’ll walk through the steps and even include prompts and images so we can inspect things as we go!

PS: Adding –stylize 1000 and changing the prompt to something completely different is REALLY fun to play with!

(click the prompt to copy it!)

/imagine prompt: a cat --style cXnsD1zCaIieSBrE

(click to view)

Breaking MidJourney with 50MP

Any mad scientists here? For the past few days I have been running the same prompt over and over in an attempt to break MidJourney – and I finally succeeded… well… kinda. I created a 50 megapixel image (16,320 x 3,072px) and then MidJourney said, “Nah man, you done!” 🤣

The new pan feature is really amazing, but I was curious about its limits (and learned some things in the process). I came up with a prompt to give some cool psychedelic sci-fi output, then added “–ar 1:8” in order to make a really tall canvas. Then I simply kept panning to the right and changing colors in my prompt, nothing else. The cool thing about panning is that it keeps your resolution, but adds more with the new generation. If you zoom out or anything, it downsizes the image again, so you have to only use pan.

Anyway, after a bunch of pans the grid images stopped displaying. Maybe a size limit on Discord or something, but instead of an image I was given a URL link – that didn’t work either, lol. So to even see the images I would upscale all four, then they showed. Finally, after 3 days of prompting, I hit the wall. After reaching over 16 thousand pixels it fails every single time. I’m sure the devs are aware of a file or pixel limit on their site or Discord or something, but I got it. I also found that if I compare an image early in the process to late, the early one actually has more detail in shadows, whereas the later images are underexposed. This shouldn’t happen, but I’m thinking it is from saving an image over and over again, losing quality in the process. Interesting. So technically I could composite the old with the new and get a  great image. Okay, I’ll shut up now. Back to my mad scientist AI experiments!

(click the prompt to copy it!)

/imagine prompt: blue color::1.5 photography of a fantastical city of crystal and holographic light, on the edge of a rocky cliff, led lights, foggy, volumetric lighting, bright beams of light, insanely detailed, gothic architecture, ultra-wide angle view --ar 1:8

(click to view)

Mixing Different Versions

You can mix two different versions of MidJourney with the new Vary Region feature! Want to give new life to an old creation? Just find an old image in your gallery, copy the job ID, then use the /show command to bring it back in Discord – then you have the option to use Vary Region (aka inpainting) to add V5.2 on parts of the image.

I started using MidJourney while it was in V2, over a year ago. I made this beach scene during my trial and was blown away at the creativity (and immediately subscribed, lol). Well, I brought it back to life by (1) using Zoom Out then (2) using Vary Region to add multiple elements, and finally (3) using Custom Zoom to change the aspect ratio to 16:9… If you want to then mix the “themes” of the image fully into V5.2 just make full variations! Kind of a fun way to play with really old ideas.

(click to view)

InPainting, aka Vary Region

When I am wrong I admit to it, and guess what – I was wrong. The royal we have been wanting and begging MidJourney to add inpainting for what feels like eons. As many of you know, I am a HUGE advocate and word-spreader for MidJourney BUT what I also did is constantly downtalk them running on Discord. If you dig around the internet you’ll likely find tons of comments from me saying they won’t beat the competition without leaving Discord behind. Little did I know they were working directly WITH discord to get an inpainting interface coded. It’s here, and it’s freakin’ amazing.

I am the lead editor for an awesome media company that does everything from commercials to events, and luxury real estate is a large focus. At Keen Eye Marketing I edit videos day in and day out. I wanted to make an image about this using AI.

What I love about AI is the unexpected outcomes. First, I used a simple prompt to get screenshots, but one was different. The house on the monitor went out into the exterior. I then used Region Remix to add a mouse and keyboard, then expanded the canvas with pan. Though I was limited now, I brought it into Photoshop, fixed a few things then added more sky above. In no time I had a really engaging image (though some things might look a little funky!).

(click the prompt to copy it!)

/imagine prompt: adobe premiere video editing interface sliders buttons --ar 16:9

(click to view)

One Year Later

I just saw this pop up in my memories and it gave me a good feeling. A year ago today I released the first version of Prompter (the prompt-building spreadsheet) for free just to see if people would use it… I never thought that a year later it would get so much attention, tons of downloads, multiple updates, and then it’s own website with notes. Not to mention all the awesome support and donations. Ya’ll rock. 🪨 A new Prompter for V5.2 is now avalable.

As for me? I’m taking a break. I recently moved to Rhode Island and decided to pick up my camera again. Hiking, OneWheeling, exploring, and editing pictures has me feeling great… I’m sure I’ll get back to AI soon enough, but for now – I hope you all keep making fun stuff for me to look at! ✌️

(click to view)

Corgi Van Gogh

We were out shopping and my girl saw some wall art of a corgi with a Van Gogh look. She wanted it, but I said “Wait, let’s just use MidJourney and Photoshop to make one that looks like Willow instead!” (our little fluffy tri-color corgi pup).

I rolled a few simple prompts (sometimes they are all you need) in different versions and styles and got one that was really close, and the Van Gogh brush strokes were really cool. I then upscaled the image in Visions Of Chaos, then used layers and generative fill in Adobe Photoshop to make it look more like her, with a little goofy eye and black stripe down her snout. I sent it off to print as a stretched canvas, and we are both (well, all three of us) incredibly happy with the new decor!

(click the prompt to copy it!)

/imagine prompt: a black tricolor corgi in the style of Starry Night by Vincent van Gogh --v 5.1 --style raw --ar 1:1

(click to view)

Blog Search

Blog Categories

Blog Archives

Be Careful Out There

Here is a very different kind of blog post. A warning, or something like that. We need to talk...

One Year Later

I just saw this pop up in my memories and it gave me a good feeling. A year ago today I released...

Corgi Van Gogh

We were out shopping and my girl saw some wall art of a corgi with a Van Gogh look. She wanted it,...

Animation Progress

Hey, been a bit quiet lately. I’m busy working on an animation project using MidJourney and After...

Pin It on Pinterest

Share This