The twunk with the flower tattoos
The twunk with the flower tattoos. The title of this set is obviously a play on the book titled “The Girl with the Dragon Tattoos,” which was subsequently made into a movie. I like to play with this name because it’s one of my favorite movies, with characters that I relate to. Thrillers are also very much of my genre.
If you’re familiar with my works, then you’d know that I love flowers, and I like pairing flowers with men. Having a contrast of hard and soft is a recurring theme in my art. I didn’t even realize this until I made reels of my works and saw the pattern.
The funny thing is — when I had that realization and subsequently wrote about it, Brawny would tell me that he saw it all along. I guess it’s true that we are blind to our own patterns.
Stable Diffusion Checkpoint Models
In Stable Diffusion, you have the power to influence the style of the artwork using prompts alone. However, for optimal results, it is recommended to use a model specifically designed for the task.
I rely on a set of checkpoint models created by Scratchproof, aka @virilevision on Instagram. If you’ve ever worked with gay AI art, you’ve likely encountered Scratchproof’s checkpoint models. He is responsible for all the Virile models, which were deliberately crafted with a male focus, making them ideal for creating gay AI art.
For this particular image set, I followed the same workflow as I did with the “Tattooed gym bro going on a date” set. I processed a photo of @BenWedgie holding a bouquet of flowers and used the same text prompt to generate a series of images using different checkpoint models published by Scratchproof.
Here are the checkpoint models I used and their respective styles:
1-2. Virile Fantasy v1.1. This model was primarily created to emulate the painterly style found in fantasy art. It can also be useful for creating paintings.
3-4. Virile Motion v1. This model closely resembles the 3D cartoon style commonly seen in Pixar movies.
5. Virile Reality 3 Beta 3. This model was designed to produce photographic works. However, with the right prompts, you can achieve results that are nearly indistinguishable from photographs.
6-7. Virile Fusion v2. This model generates photorealistic illustrations with smooth contours and vibrant colors.
Tech
- Stable Diffusion txt2img with control nets
- Steps: 50, Sampler: DPM++ 2M Karras, CFG scale: 7, Size: 512x512
- Model hash: 59bcb95ef6, Model: virileAnimation_v10
- VAE hash: 63aeecb90f, VAE: vae-ft-mse-840000-ema-pruned.safetensors
- Hires upscale: 3, steps: 50, upscaler: 8x_NMKD-Superscale_150000_G, Denoising: 0.5, 1536x1536
- ADetailer model: face_yolov8n.pt, steps: 50, ControlNet module: tile_resample
- ControlNet 0: depth_midas, Weight: 1, Guidance Start: 0, Guidance End
- Automatic1111: v1.6.0-2-g4afaaf8a
- Gigapixel HQ 4x, 6144x6144
- Adobe Photoshop AI for outpainting
- Adobe Lightroom
