Stable Diffusion

Text-2-image

Stable Diffusion 1.5 & Stable Diffusion XL

With this node you can generate images of anything. Available with more than 80 models

Quick menu: click on any section to quickly go to it

Key features

Step by step tutorial

Difference between Stable Diffusion 1.5 and Stable Diffusion XL

Parameters description

Recommendations for best results

Key features

Realistic and artistic images
Capable of generating in the numerous style depending on the prompt. High variability
Stable Diffusion 1.5 and XL
Available with 2 state-of-the-art models: Stable Diffusion 1.5 with option to train the model and 80+ available styles, and Stable Diffusion XL for better quality
Generate in up to 1524x1524. Upscaling up to 4k
Stable Diffusion XL gives incredible results at higher resolution, whereas Stable Diffusion 1.5 provides better quality at lower resolution
Wide choice of custom models
With custom checkpoints (Styles or models) you can get the desired images in various styles and photorealism.
More than 80 styles available with most popular Dreamshaper and Juggernaut models. Includes SFW and NSFW models
Fast generations
Starting from 10 seconds you can get the desired images in batch generation up to 20 images per generation (depending on your subscription)
Basic and advanced settings
Whether you're new to AI art or an experienced user, you will find all the necessary parameters and high customizability for your use case

How to use the node

Click on each step to see the corresponded image as an example and tutorial to follow

1. Type your idea
Neural network creates images based on your text. This text is called prompt.
The more precise your prompt is, the better the results will be
2. Choose Style
In this node you can generate with more than 80 styles, each style is an improved Stable Diffusion model, which was trained to give better results or achieve a specific style.
If you don't know which Model or Style to choose, try Deliberate, Reliberate or DreamShaper
3. Launch
Press the green button to launch the AI magic process.
You can download the results immediately or upscale it to achieve higher resolution.
If you don't like the result, press Start button again - it will change Seed automatically and give you other results with the same prompt
4.Optional: change the Style
By changing the Style you can significantly improve or change the result. Try out different styles to find the best combination for your prompt

Difference between Stable Diffusion 1.5 and Stable Diffusion XL

Stable Diffusion 1.5

More models

SD 1.5 has more than 80+ available custom checkpoints, meaning more variability

Trained Models

Possibility to use trained AI models

Recommended resolution

512x512

Stable Diffusion XL

Higher quality

By default it generates 1024x1024, meaning less artifacts and better images

Easier prompts

Works way better with shorter prompts than SD 1.5

Recommended resolution

1024x1024

Need inspiration for prompts?

Check out our prompt collection!

Explore prompts

Parameters

Generally as a basic user you would only need to type prompt, negative prompt, set up resolution if needed. However, you might need to change other settings, so here are the descriptions of what each parameter does.

My models

A dropdown list of your DreamBooth models
Available in Stable Diffusion 1.5

When turned on, it will use your trained model from AI Photoshoot node or Train Panel.

In order to use it you have to write down the subject in the Text prompt you trained it on. By choosing a model, the subject will be placed automatically to your prompt, but if you erase it, you would have to type it again

Important note: you can NOT use DreamBooth model and Style (custom checkpoint) at the same time

Styles

A dropdown list of custom checkpoints, custom models or Styles
Available in Stable Diffusion 1.5 and Stable Diffusion XL

Currently we are offering a big number of available checkpoints (80+ at the moment)

If you have any suggestions, which custom models you want to add to the Phygital+, write us in the Discord

Steps

The range: from 1 to 200
Available in Stable Diffusion 1.5 and Stable Diffusion XL

It defines how detailed and close to the prompt your image will be. The higher number of steps is, the longer it takes to generate one image.
Good results can be achieved in 30-50 steps

Usually the more steps, the better, but we recommend to experiment with 50 or 100 steps, as it's the most optimal time spent and quality correlation

Seed

The range: from 1 to 4294967296
Available in Stable Diffusion 1.5 and Stable Diffusion XL

The generation starts from a particular image (or noise). The same seed number will give you similar results each time, whereas by changing it to another number may dramatically influence the final output.

We recommend to use different seed numbers if you're experimenting, and as soon as you find the perfect result

Width and height

Range: from 512 to 1024 (in SD 1.5) and to 1524 (SD XL)
Available in Stable Diffusion 1.5 and Stable Diffusion XL

It defines the width (x) and height (y) of the final image.

We recommend to use
horizontal orientation for landscapes and environments,
vertical for characters and creatures

No (optional)

Available in Stable Diffusion 1.5 and Stable Diffusion XL

Negative prompt. It tells neural network what it should NOT generate

Scale uncond (CFG Scale)

The range: from 0 to 100
Available in Stable Diffusion 1.5 and Stable Diffusion XL

It defines how much AI follows the prompt. It also Impacts on image contrast.
The higher the number is, the more vibrant and contrasted the image will be.

We recommend to keep it in the range of 7,5 to 15

Eta (Guidance Rescale)

The range: from 0 to 1
Available in Stable Diffusion 1.5 and Stable Diffusion XL

Impacts on image depth and color range

We recommend to keep it around 0.8 so that the picture is more balanced.

Samplers

There are multiple samplers
Available in Stable Diffusion 1.5 and Stable Diffusion XL

It defines how AI will generate the image, that's why by choosing a different sampler you can get different results, but some of them don't affect the image that much.

We recommend to change it only if you're an experienced user

Start image skip

Range: from 0.05 to 0.95
Available in Stable Diffusion 1.5 and Stable Diffusion XL (img2img)

It defines how close to the Start image the final image will be.
The higher the number is, the more different from the Start image a generated artwork will be.

We recommend to keep it between 0,5-0,6 if you use Start image

Tiling X axis / Tiling Y axis

Available in Stable Diffusion 1.5
Makes the image seamless and tileable

Perfect for creating textures

Artistic Mode

Available in Stable Diffusion 1.5
It turns any generation into the artistic masterpiece.

Currently available at 4 styles:
• General Artstic mode: for anything from objects to personas
• Portrait: for personas and any humanoid-like creatures
• Full-body: for full body height images of humanoid-like creatures
• Landscapes: for cinematic scenes

Denoising end

SD XL parameter
The range: from 0 to 1

If set to 1, the refiner is not used in generation.
The less the number, the more % of generation time will be split between the general model and refiner

Number of images

Available in Stable Diffusion 1.5 and Stable Diffusion XL

Determines how many images will be generated per one node launch.
The highest number is 20 for SD 1.5 for Plus subscribers

Getting bad results? Try one of these!

Use different Style

Some Styles perform exceptionally well with short prompts. Try these Styles:
DreamShaper, Deliberate, Juggernaut

Change Resolution

A square image doesn't always give a good result. Try adjusting size for the image for:
• characters: vertical orientation (e.g, 512x768)
• landscapes: horizontal orientation (e.g., 768x512)

Try negative prompt

Fill in 'No (optional)' text field.
Write down what you do NOT want to generate. It can help a lot with results

Try adding these words: weird, ugly, deformed, poor quality, blurred