Sign In

Lz16K+Optical

323
2k
18
Verified:
SafeTensor
Type
Checkpoint Trained
Stats
1,407
Reviews
Published
Sep 24, 2023
Base Model
SD 1.5
Hash
AutoV2
C6C7EAE73E
default creator card background decoration
Lunzi_'s Avatar
Lunzi_

16K+optical image realistic portrait photography

Introduction: The lz16k+model mainly optimizes its ability to generate excellent lighting and realistic portraits without relying on prompt words. It can also optimize facial details of characters while generating high-definition magnification, simulating very realistic photography effects. Simple tags can also generate good portraits, and there have been many optimizations in composition. The model recommends portraits or busts, Combined with high-definition restoration and the afterdetailer plugin, the entire body is also very good. In addition to portraits, it has good output for interior design, architecture, daily necessities, and more. We hope that everyone can explore the potential of this model for more testing!

Parameter recommendation:

TAG:

-(EOS R8,50mm, F1.2,8K, RAW photo: 1.2), FUJIFILM X-T4

-Real skin texture, Realistic photo, best quality, masterpiece

-Studio lighting, High resolution scan, physically based rendering, Cinematic Lighting

-AWARD Winning Work, depth of field, (low saturation: 1.3),

(Professional photography, please make your own adjustments for reference only)

Size: 512x768, 1024 * 1536

Amplification algorithm: 4x UltraSharp/8x_ NMKD Superscale_ 150000_ G

Magnification factor: 2.0-2.5

Sampler: DPM++series

Iteration steps: 30-150

CFG Scale: 7

VAE: built-in

Do not enable facial repair - ADetailer plugin recommendation: face_ Yolov8n. pt, face_ Yolov8s.pt

If you need to generate higher quality works, it is recommended to use LRC and PS for post processing!

Creative inspiration: The current live action series of large models has undergone significant changes from the initial skin and face to light and shadow, but most of them have outstanding effects at low resolutions. Therefore, we hope to create a model that still has excellent details in light and shadow imaging after high-definition magnification. Currently, there are still many details that need to be improved. Stay tuned.

If you have task questions, you can leave comments and comments!