HOME | DD

Description
Do you remember that first kiss, when your lips first touch, that anticipation, that excitement, that nervousness, that feeling of being alive?

Web: gymdreams8.com/blog/first-kiss…
Instagram
Set 1: www.instagram.com/p/CumQ06jLRB…
Set 2: www.instagram.com/p/CumO0aKLab…
Set 3: www.instagram.com/p/CumOZQLrgV…
Twitter
Set 1: twitter.com/GymDreams8/status/…
Set 3: twitter.com/GymDreams8/status/…

These were made for my Kiss series, but are also test renders for Cyber Realistic, a checkpoint model for rendering photorealistic works in Stable Diffusion that a friend had recommended to me. Full post is quite long but I’ve provided a summary here. You can read the full post on my site at gymdreams8.com - link in bio.

When I test checkpoint models, I look for:

- How it renders skin and skin tone
- How skin looks under direct vs indirect (soft) lights
- How well does it render “hairy chest” — does it have volume or does it look flat and close to the surface
- How well does it render facial hair — smooth, clean shaven, stubble and full beard

I then rendered men of different ages and body types (slim to muscular) to see the range of muscle mass it can produce.

For these, I combined all tests into one, by using Regional Prompter:

- 45yo muscular man, beard, hairy chest
- 30yo athletic man, clean shaven, smooth body

I then added the mood that I want to portray: the delicateness of that first kiss, the fog for romance, a nice bokeh to visually separate the foreground with the background, and rain to add some texture in the air.

The results from Stable Diffusion txt2img were already beautiful, but as with most AI renders, some of the hands were problematic. So I took some of these images into Photoshop, and used the Generative layers to fix them.

Overall, I am quite happy with these results, and will likely use this model again for future renders.

25 steps, DPM++ SDE Karras, CFG 5, Cyber Realistic v3.1, Denoising 0.5, ADetailer: mediapipe_face_full, Regional Prompter, Hires: Latent 2x 20 steps, Gigapixel HQ 2x, Lightroom, Photoshop AI.

Description
Do you remember that first kiss, when your lips first touch, that anticipation, that excitement, that nervousness, that feeling of being alive?

Web: gymdreams8.com/blog/first-kiss…
Instagram
Set 1: www.instagram.com/p/CumQ06jLRB…
Set 2: www.instagram.com/p/CumO0aKLab…
Set 3: www.instagram.com/p/CumOZQLrgV…
Twitter
Set 1: twitter.com/GymDreams8/status/…
Set 3: twitter.com/GymDreams8/status/…

These were made for my Kiss series, but are also test renders for Cyber Realistic, a checkpoint model for rendering photorealistic works in Stable Diffusion that a friend had recommended to me. Full post is quite long but I’ve provided a summary here. You can read the full post on my site at gymdreams8.com - link in bio.

When I test checkpoint models, I look for:

- How it renders skin and skin tone
- How skin looks under direct vs indirect (soft) lights
- How well does it render “hairy chest” — does it have volume or does it look flat and close to the surface
- How well does it render facial hair — smooth, clean shaven, stubble and full beard

I then rendered men of different ages and body types (slim to muscular) to see the range of muscle mass it can produce.

For these, I combined all tests into one, by using Regional Prompter:

- 45yo muscular man, beard, hairy chest
- 30yo athletic man, clean shaven, smooth body

I then added the mood that I want to portray: the delicateness of that first kiss, the fog for romance, a nice bokeh to visually separate the foreground with the background, and rain to add some texture in the air.

The results from Stable Diffusion txt2img were already beautiful, but as with most AI renders, some of the hands were problematic. So I took some of these images into Photoshop, and used the Generative layers to fix them.

Overall, I am quite happy with these results, and will likely use this model again for future renders.

25 steps, DPM++ SDE Karras, CFG 5, Cyber Realistic v3.1, Denoising 0.5, ADetailer: mediapipe_face_full, Regional Prompter, Hires: Latent 2x 20 steps, Gigapixel HQ 2x, Lightroom, Photoshop AI.