Z-Image Power Nodes is a collection of nodes designed specifically for the Z-Image / Z-Image Turbo model. They are based on some ideas and discoveries I made while developing the Amazing Z-Image Workflow.
This page includes example workflows that utilize these nodes. The nodes can be installed via ComfyUI-Manager or manually from the GitHub repository. For more information, please visit:
These are the nodes currently included in the project:
⚡Z-Sampler Turbo ^G2
A specialized sampler designed for Z-Image Turbo that achieves sufficient quality to eliminate the need for further post-processing.⚡Style Prompt Encoder
Applies a selected visual styles to your prompt and encodes both of them using a text-encoder model (clip).⚡Style String Injector
Seamlessly integrates a chosen style into your prompt text. It accepts a string as input and modifies it based on the selected style.⚡My Top-10 Styles
Allows you to create a list of favorite styles for quick selection of your most used ones.⚡VAE Encode (for Soft Inpainting)
Encodes images into a latent representation, embedding the mask that indicates where inpainting will be applied.⚡Save Image
Saves generated images with the option to embed CivitAI-compatible metadata, making it easy to share generation parameters through that platform.⚡Empty Z-Image Latent Image
Creates an empty latent image of the appropriate size for Z-Image, selecting aspect ratio, scale, and orientation.
Description
This release introduces the second generation of "Z-Sampler Turbo", with the internal code completely rewritten from scratch. It follows the same core principle as the first generation (3 stages with their respective sigma sets) but adds new functionality and, more importantly, better parameterization of the internal sampling process.
Thanks to this rewrite, I've been able to implement the following controls:
* Intensity: adjusts how intense or "washed out" the details in the final image appear.
* Intensity Bias: allows you to adjust the "brightness" and details produced by extreme intensity values.
* Turbo Creativity: aims to bring more compositional variety to the image without altering its style or color tone.
FAQ
Comments (7)
Really cool what this style encoder does with simple prompts. Quite fun to use. Looking forward to try all the styles with different prompts.
Glad you're enjoying it! In this version, I've also included a workflow 'z-image-turbo__t2i_double_trouble' that lets you mix two different styles, giving way more combinations to play around with.
Excellent work, thank you very much, it keeps getting better.
Thank you so much!
What does the new experimental Spectral Tilt feature do?
What is the meaning of:
spectral_tilt_start
spectral_tilt_end
spectral_tilt_sharpness
And what is the difference between:
stage3
stages23
stages123
stages12X
I tested with fixed seeds and could not notice any visible difference.
Could you explain the specific usage? For example, the values for the scene and the portrait respectively. Using the default values seems to be a waste of this new testing feature.
The Spectral Tilt features in the "Z-Sampler Turbo 2.1" node are still super experimental, which is probably why you're not seeing changes yet. Under the hood, it basically tweaks the noise frequency distribution injected at each step during Euler Ancestral sampling.
The different stage options just pick where that tweaking happens within the sampler's three stages. The early stages (1 or 2) subtly nudge the overall composition, while the later stages (2 and 3) focus on the tiny micro-details. Since it's still a work in progress, it might fail in some situations. I've noticed it can nicely polish up the final details or at least give you a cleaner alternative version of the same image. These experimental nodes might keep changing while I continue developing and testing them, but in one or two days I'll try to release a more polished version. Thanks for the feedback.
I create 1200 * 1920 pictures within 30 seconds and quality is so good !!! Thanl you one million times !!!! 💕