Humiliation Level: Maximum

I no longer specify every time that my work is AI-generated, as it became a bit tedious to paste the same paragraph repeatedly. Here’s a reminder:

Images are generated using NovelAI with the Opus Tier. This paid tier allows unlimited and fast generation, enabling me to create a lot of variations and experiment basically everything I can think of.

However, to produce a really nice image, prompts and rolls are not enough. I use NovelAI’s advanced functions like Vibes Transfer, Character Reference, and Inpainting. Inpainting is particularly useful, as it allows to correct each part of an image, and I use it A LOT.

These advanced functions make prompt sharing almost since you cannot regenerate the exact same image using the seed and prompt alone. If you’re interested, you can check my Rentry page, where I explain my workflow in more detail. I also plan to add a tag section soon, listing my most frequently used tags.

All text and titles are written by myself. I only use a language model for grammar correction, since I’m not a native English speaker, and for Japanese translations when I post on Pixiv.

Animated content is generated using Tensor Art with the Pro Tier, which gives 300 credits per day. That's enough to try multiple shots and keep only the best one. Unlike NovelAI, I dont have a lot of freedom where i can use my "skills" to drive AI. It mostly comes down to a good prompt and hoping the AI generates a coherent animation.

Sometimes I do video editing using DaVinci Resolve. It’s not too complex but you need to understand the basics. Mostly, I use it to link multiple shots generated from Tensor Art into a single sequence with smooth transitions. Finally, I use Gifski to convert the MP4 into a GIF.

Again, you can check my Rentry page if you want more details!