Click [HERE] to download this Blender model file. Click the image above to see the original 1920 x 1080 resolution image.
This model was pre-rigged, meaning if we select a bone and move it, a corresponding part of the mesh will follow along.
Let’s practice. Click this bone to select it, and type [R] to rotate it.
That worked well. Let’s try the Right Arm as well.
Press [Control + Alt + Numpad Zero] together to Move Active Camera to View.
If your keyboard doesn’t have Numpads, you can tell Blender to treat the standard number keys as Numpad keys by checking Emulate Numpad under [Edit -> Preferences -> Input].
Press [Shift + `] to activate the ‘Walk Navigation’ mode.
While in ‘Walk Navigation’ mode:
- Move the mouse to pan the view.
- Move the camera forward/backward W, S.
- Strafe left/right A, D.
- Move up and down Q, E.
Please press [F12] to render the scene, and save the image.
Please copy and paste the following prompt details.
Prompt: (red eyes), Photograph of scary and sinister metal skull robot with transparent head, darkness, night city background, stars, detailed circuitry and wires visible inside the transparent armor and face, post processing, post production, detailed and intricate, tone mapping, hyper realistic, photorealistic, ultra photoreal, ultra-detailed, intricate details, 8k, super detailed, full color, surrealism, full detail, ray tracing, best quality, extremely detailed.
Negative Prompt: (worst quality, low quality, normal quality:1.8)
Import Blender’s render image file into ControlNet by clicking ‘Click to Upload’ button.
Once the image loads, it would look like this.
Please use the following ControlNet options under Control Type:
1. Depth.
2. Preprocessor set to None.
3. Mode set to sai_xl_depth_256lora.
‘sai_xl_depth_256lora’ file can be downloaded here.
https://huggingface.co/lllyasviel/sd_control_collection/blob/7cf256327b341fedc82e00b0d7fb5481ad693210/sai_xl_depth_256lora.safetensors
Click [Generate] to generate a new image, and click this picture icon to send the result to img2img.
Please use the following options:
- Width: 960.
- Height: 540.
- Denoising strength: 0.9.
Click [Generate] to generate a new image, and send the result to img2img to generate a new image.
Please repeat the process to generate the best looking result.
This is the end of the tutorial. 🙂