MeTi Z-Image Workflow (ControlNet+UpScale)

Hello everyone 🌿

At first, I was using this workflow:
Z-Image Turbo Workflow with SageAttention (Multi-GPU)
(link: https://civitai.com/models/2170120/z-image-turbo-workflow-with-sageattetion-multigpu)

I kept the description and recommendations from the original workflow, but almost everything else has been changed, edited, and rebuilt to better suit my own workflow style.
I’m not a professional or an expert — just an ordinary user who wanted to share this edited version in case it might be useful to someone ✌️💛

Notes & Improvements Added

• I fixed some issues related to latent size and added manual latent select for T2i.

• I added clearer notes about resizing input images.

• The spot where you can use ImageScaleToTotalPixels instead of Resize Image v2 is marked with arrows and I added notes there to make it easy to find.

• In general, the important things that should be checked are highlighted in yellow inside the workflow.

• This new file replaces the old one.

Workflow Structure

The workflow is organized into four main sections:

1. Text to Image – Generate an image directly from text

2. Image to Image – Rebuild or modify an image using another image as input

3. ControlNet – Provides precise control over structure, pose, depth, lines, etc.

4. Upscale – Improves the final resolution and quality of the output

🔸 Important Note:
When using the workflow, use either Text to Image or Image to Image, along with ControlNet and Upscale.
Do not activate both Text to Image and Image to Image at the same time, as this may cause conflicts or undesired results.

• I got good results using the Euler sampler with Beta.

• Recommended steps: between 8 and 10.

• Tested on RTX 3080 GPU

• These settings worked well in my tests, but feel free to adjust based on your own needs.

Special Nodes Added

To make the workflow more practical and user-friendly, several custom nodes were added:

• Lora Manager
  → For easy LoRA handling
  → Includes automatic LoRA tag insertion
  → Makes selecting, activating, and combining LoRAs much simpler

• Camera Control
  → Controls camera angles
  → Automatically adds camera parameters to the prompt

• Color Adjustment Nodes
  → For color correction, tone adjustment, lighting balance, and overall color harmony

• Bookmarks
  → For faster access and better organization of workflow sections

Final Note

If you notice any issues or problems with the workflow, I sincerely apologize 🙏
I made this version in a limited amount of time between other tasks, but I wanted to share it with you anyway.