#38547 - safe, ai composition, editor:marestare, generator:chromaxlmix_v4sorbet, generator:stable diffusion, generator:zoinksnoob, prompter:marestare, twilight sparkle, pony, unicorn, g4, cloud, daisy (flower), detailed, ears up, female, field, forest, full body, hooves, horn, implied rainbow dash, looking up, mare, mountain, offscreen character, open mouth, outdoors, rainbow trail, scenery porn, shading, sitting, sitting on grass, sky, sky background, solo, sonic rainboom, stunned, tail, underhoof, unicorn twilight, wallpaper

Rainbow and Twilight are hiking in the mountains, enjoying their time together 🥰. The atmosphere is so clear and calm that Rainbow Dash decides to break the silence and astonish her friend with a Sonic Rainboom 🌈.

I’ve been studying ComfyUI lately, and it’s a piece that I used as my learning sandbox. I’ve also been trying the ChromaXL model, which I’d say is almost the same as Zoinksnoob, but it requires fewer sampling steps for generation.

I decided to add a Sonic Rainboom to the scene to make it a bit more alive. It was rather difficult to blend into the picture. Although I still think it could be done better, it’s good enough already.

Great thanks Echo159 for huge help with sharing their knowledge about AI art. I’ve been able to find some mistakes in my usage of Zoinksnoob and improve the quality of the generated images with it.

I used both ChromaXl and Zoinksnoob for this one. Chroma was used to generate the base image, and Zoinks was used for upscaling and inpainting.

My X (Twitter): SFW, NSFW

Syntax quick reference: **bold** *italic* ||hide text|| `code` __underline__ ~~strike~~ ^{^sup^} _~sub~

Detailed syntax guide

---

I've been studying ComfyUI lately, and it's a piece that I used as my learning sandbox. I've also been trying the ChromaXL model, which I'd say is almost the same as Zoinksnoob, but it requires fewer sampling steps for generation.

I decided to add a Sonic Rainboom to the scene to make it a bit more alive. It was rather difficult to blend into the picture. Although I still think it could be done better, it's good enough already.

Great thanks [Echo159](https://tantabus.ai/profiles/Echo159) for huge help with sharing their knowledge about AI art. I've been able to find some mistakes in my usage of Zoinksnoob and improve the quality of the generated images with it.

I used both ChromaXl and Zoinksnoob for this one. Chroma was used to generate the base image, and Zoinks was used for upscaling and inpainting.

My X (Twitter): [SFW](https://x.com/mare_stare), [NSFW](https://x.com/mare_stare_nsfw)

https://derpibooru.org/images/3536258

https://e6ai.net/posts/87933

https://x.com/mare_stare/status/1883655800733983055

Tantabus

BBCode

Comments