Basics of image generation
Soulkyn offers it’s users enhanced capabilities for image generation.
This page will teach you the basics of what our image generation has to offer.
Explaining the UI
Let’s get familiar with our image generation interface!
| UI element | Function |
|---|---|
 |
This button lets you choose image model you want to use to make your pictures. Simple explanations with examples are provided below. |
 |
This button shows advanced settings which include LoRAs and other functions. Because of how complicated it is the explanation will be provided in the Advanced Tutorial for Images. |
 |
this three buttons let you choose how many pictures do you want to make. |
 |
This button toggles between SFW and NSFW prompts. Please remember that those are here to help you and are self explanatory. You don’t need to choose NSFW prompt to make one. It’s just easier with them. |
| Prompt that let you choose the overall look of the image. They only steer the AI in the specified direction it does not force the chosen prompt. | |
 |
This button lets you browse through your background presets. Those are created automatically whenever you make a new Kyn. This button also has a search function! Just highlight everything (CTRL+A) and write which background you want to find. |
 |
This button lets you browse through your clothing presets. Those are created automatically whenever you make a new Kyn. This button also has a search function! Just highlight everything (CTRL+A) and write which clothing you want to find. |
 |
This button lets you add a new preset. Please remember that the minimum of tags for a preset is 3 (three). |
| Here you can input your image prompt! Just describe what you want in simple terms and it should appear! | |
 |
This lets you save and load basic image prompts. |
 |
This button starts image generation! |
After you press the continue button you should be able to see generation UI
| UI element | Function |
|---|---|
 |
This button checks every picture to be deleted. |
 |
This button confirms all your choices |
  |
Button that sets a particular picture as an avatar picture for your Kyn. NSFW pictures aren’t allowed to be avatars |
  |
This button sets an image to be deleted after pressing continue |
Image models
Anime original
Basic most versatile anime model.
Anime Alpha
This anime model is very rigid and harder to break.
Anime hyper
An anime model with more detailed shading
Realistic original
This realistic model is best used for more grounded and realistic pictures
Realistic freedom
This realistic model lets you easier make pictures related to fantasy
ZiT Models
ZiT models by default are realistic but with tags like anime style or describing that it’s an anime illustration you can easily change the artstyle
ZIT - All
A versatile model trained in-house. Unlike SD models this accepts bot tags and descriptions.
For anime I had to use lora at low value
ZIT-Photo
A versatile model trained in-house. Skewed towards photos. Unlike SD models this accepts bot tags and descriptions.
ZiT-Real
A versatile model trained in-house. Skewed towards realistic. Unlike SD models this accepts bot tags and descriptions.
Your first image prompt
Before you learn how to make pictures you need to understand how to write so AI can understand you.
First of all:
Soulkyn currently has two vastly different image model bases: SD and ZiT.
In case of SD the way you need to write your prompt is tag/keyword based.
What are keywords
Keywords (often called tokens or tags) are the bridge between your imagination and the AI’s “brain.” Think of them as specific instructions that tell the model which parts of its vast library of data to pull from.
How They Work
When you type a prompt, the AI doesn’t see a sentence; it sees a collection of tags. It uses these to narrow down the look, feel, and subject of the image.
- Subject Tags: Tell the AI what to draw (e.g.,
cat,cyberpunk city,astronaut). - Style Tags: Tell the AI how it should look (e.g.,
oil painting,3D render,minimalist). - Quality Tags: Help the AI refine the details (e.g.,
highly detailed,8k resolution,masterpiece).
This means to make character sit on a chair in living room you need to use special wording ex: Sitting, on chair, living room
the resulting picture:
Please remember to put a comma after every keyword that you use! They help ai differentiate between meanings of the words.
If you decide to use a ZiT model you can use both keywords and descriptions to create your pictures. This is the same as for example Midjourney or other sites where you can create pictures.
Beside that you can use weights to tell AI how much you want certain thing! Basic weights can be applied with + and -
Most common weights:
+ +
+
-
--
Actual weight system is really complicated and can be used for many fantastic things! More on that you will find in the Advanced Tutorial for Image Generation.
Now let’s make something that you can use for your first Kyn!
Let’s begin with choosing preset: #Portrait will make do for now
Now let’s make a proper prompt for the avatar!
Let’s say we want to make her avatar picture to be a portrait picture where character is standing alone and smiling while winking. The background will be blurry.
Image prompt: Solo++, Close++, Portrait++, Smiling-, Winking+, Blurry+ background++,
The result:
There you go! Your first avatar for your Kyn! You can now go to the advanced tutorial if you want to really get into image generation!