I discovered a secret to unlocking unlimited variations in Midjourney. I used Vary Region inpainting differently than most people. I'll show you a peak at my new workflow: [img:UxqFgJDmn]

People have mentioned that it's possible to improve an entire image using Vary Region and masking the whole picture. I started to do that, but I left the full prompt & clicked generate. *Pay attention to the original image to see the huge difference compared to the results. [img:vid-vLvO2ir9S]

I got a new grid of results and re-rolled the grid to get more and more variations. But you can see the new grids have elements of the original, but each new result had completely different variations. Starting image: [img:Gqyl58OT9]

I realize the results are not my typical AI art style, but why would I expect them to be so different from the original? Re-roll 1: [img:0JMTuHdc7]

Interestingly, you can see similarities for each result, but expressed differently. Re-roll 2: [img:shQ6XGado]

Each grid includes 1 image with yellow effects, and red hair, all with extreme or exaggerated eye makeup and bright lips. Re-roll 3 [img:Byy645qSM]

Additionally, I noticed character consistency occurs frequently. I usually have at least one character repeating in different variations on each grid result. It's an interesting way to generate interesting ideas. Try it and reply with your results 👇 [img:LbvkffeD_]

Thanks for reading. Follow @HBCoop_ to explore AI art tips and techniques with a fresh perspective.  Repost the 1st post below to share this thread: twitter.com/HBCoop_/status/1706048433398702571