Pushing the Boundaries: In the Virtual Studio this weeks Midjourney explorations


the post image above was the blend of these three images.

the middle floral in the IMAGE PROMPT above was a blend of these two images.
As you can see the blending (combining) of images can go on forever. The final results have details or textures of the previous. Perhaps one day Midjourney will give us more options on how those blends happen. More on the /blend command below.
Midjourney experiments, exploring techniques, prompts and commands.
Over the past two months, I’ve created over 3000 images using Midjourney, and it’s been a wild ride. From experimenting with photorealism to trying my hand at Claymation and watercolor styles, I’ve had a blast exploring all that this AI tool has to offer.
One of the most incredible things about Midjourney is the level of detail and realism that some users are able to achieve. It’s truly amazing to see the magical and bizarre figures and landscapes that people are creating with this tool. From what I’ve seen, most users are focused on photo-high realism, game fantasy art, 3D modeling and Manga cartoons. In fact Manga art has it’s own version of Midjourney called Niji.
Personally, I find myself drawn to art that has layered and textural effects. Very much like the art I created with printmaking techniques (see the printmaking section of my gallery. My focus for right now in AI image generation is exploring the ai-styles of linocut, lithography, and etching, and then blending those designs with digital painting and photography to perhaps one day discover something unique.
Despite my love for Midjourney, it hasn’t been without its challenges. Sometimes, the prompts can lead me in frustrating directions, and it can be a hit or miss experience. But this week, I’ve been experimenting with the /describe and /blend commands, and it’s been a game changer. /describe has helped to break down the prompts of successful images, to get to that secret sauce. And /blend allows me to select, upload and combine multiple images for stunning results.
Of course, as with any tool, there are limitations to what AI can do. The resolution is very small, but perfect for social media. To take an image to print you would have to digitally repaint it in a raster program like Procreate, (the most affordable) or to run it through a program like Topaz Gigapixel Ai (for the more serious).
Quirky image results are often entertaining & often annoying.


quirky text and logo thingys that pop up.



and they merge!

but it wasn’t the intended visual story I was trying to write for my grandkids.


the only option is to do a photo composite in procreate.
Another frustration is getting two character together in the same space without Midjourney trying to mix and merge images together. See the bunny-owl and owl-bunny. Some users have succeeded in getting mulitple characters on the same stage, but I’m not sure if they sold there souls to get there? I suspect there were some very determined efforts in remixing and blending of images. For me it’s still a mystery that I hope to solve! Midjourney is developing quickly, and maybe a simple solution is coming. Overall it’s exciting to think where Midjourney will go in the next 3 months, 6 month and 1 year.
The /blend command





Ai – abstract painting of Kouchibouguac National park




I remixed endlessly

I blend them

abstract still-life & etching images from above


The /blend command possibilies are exponential, especially when I start to consider the blending of blended images 🤯.
Note: Many times when I blended a photo with an illustration, Midjourney gave weight to the photo. I tried changing the order of the uploaded images and using –iw 2 parameter, but never had it work like I wanted. So to bypass that quirk. I decided to put 2 similar illustrations in the mix with the photo in hopes that Midjourney would give weight to the illustration, not the photo. And it worked. (Warning Midjourney doesn’t like it when you double up on the exact image). You can blend up to 5 images, so there is still lots to explore with the /blend command!
Hope you enjoyed my favorite images from this week’s explorations. Thanks for following my story with an open mind and letting me share my Midjourney experience.
Resistant to this Ai thing?
For those who aren’t embracing the world of AI art. Especially if you’re working in a commercial space, gate-keeping and resisting this technology may not be the best approach. Instead, taking some time to understand how AI works can help you differentiate your skills and stand out in the industry whether you use it or not.
The fact is, AI art is here to stay, and being open to it can offer new possibilities for your own creative process. Try it and see where it takes you? “Better the devil you know, then the one you don’t?” You can learn more on previous blog post so hope you will check them out.
HOW TO of Midjourney Program
The Midjourney program is run through the Discord server. I had to take an Udemy course by Scott Bromander to figure that out and get a basic overview of the Midjourney program. Once that is set up it’s just typing on a command prompt. If you’re of a certain age you might remember DOS. To initiate an image the command is /imagine adding the description to the prompt, it’s best to review Midjourney’s help section as I can attest that many user prompts will lead you astray, often using parameters incorrectly and that only worked with earlier versions of Midjourney.
And it is all public, you can see the work of others both on servers and in the gallery searches, including their uploaded images, and prompts. The idea is that in keeping it public you have the opportunity to deconstruct and try prompts for yourself. That said images on the Midjourney server flash by so fast that it’s chaotic and you will need to go to your Midjourney user account to find your images most of the time. There was a free access to try Midjourney which allowed 25 images to start but that has been temporarily suspended due to abuse. For now, you will need to get a paid monthly US$10.00 or $30.00/ month or discounted yearly plan. But you can cancel anytime upgrade or downgrade as needed.
With a plan, you can set up your own server to generate your images in a calm and peaceful manner. Keep in mind you are still public. Only the Stealth $60.00 per month plan allows you to keep your images private, which would be advisable if you’re working for a client. Be sure to check out the copyright, terms of service and privacy . And no funny business is allowed it’s PG13. You can get banned pretty quickly if you try to create elicit or controversial images.