Bubble 1
If you are one of the few people who have been looking at my works for a while, and you read my very detailed description of the images, where I talked about the subject matters, context of why I made the images, as well as the technical details and how I arrived at the results, then you’d know that I have always found image prompts in Midjourney to be one of its most powerful features.
That’s something that I have tried to replicate inside Stable Diffusion, but I haven’t been able to do it successfully until a couple of days ago. This is a major breakthrough for me because now I have just unlocked unlimited image generation potential.
This blue set is the first ones where I saw that I have just landed on a gold mine. The craziest thing might be the fact that these these forms may be wildly different from each other — an octopus, a dolphin, a splash — but in fact share very similar prompt, made possible because of my new workflow.
On Instagram, I posted some images with mosaics because I am not entirely sure if they are ok or not. I only intend to make safe-for-work images, but some of these early experiments resulted in images were done before I put any specific negative prompts to remove those “artifacts,” but they are so beautiful that I just wanted to share.
As my progress is still ongoing, I will label these in the steps that I produce them. This first batch will be Bubble 1, then the next iteration will be Bubble 2, etc. They are all related.
Tech: Text prompts in Midjourney v5.2 + Stable Diffusion: 25 steps, DPM++ SDE Karras, Airfuck’s Wild Mix v1. Post: Topaz Gigapixel HQ 4x, Adobe Lightroom color correction.
