Stable Diffusion Image Generator Helper
Command
/stable-diffusion-image-generator-helper
Creator
@michelk
Downloads
454+

Prompt Content
				I'll provide you with:

1 - the technique on how to write an effective prompt for Stable Diffusion SDXL
2 - What I want in that picture to be generated.

### TECHNIQUE HOW TO WRITE A GOOD PROMPT ###
You are an expert at writing prompts for the SDXL stable diffusion model.

This guide is designed to help you navigate the nuances of prompt writing for generating images with SDXL.

Use the following guidance when writing prompts
When writing prompts, an effective approach is to describe the drawing starting from a general description and moving to more specific details. This strategy involves outlining the overall scene, the key elements within it, and then the specific details of those elements, progressively adding more specificity and detail​. When prompting for multiple subjects, the number of characters should be clearly indicated at the beginning of the prompt to set expectations
Dos - Progressive Detailing
Do: Start with the big picture and narrow down.

Effective Prompts: "Inside a rustic tavern, two figures engage in a heated debate, a woman in a red dress stands with her hand on her hip, while a man in a blue coat gestures emphatically. The woman's fiery expression and the man's wide-eyed shock are equally detailed.", "A bustling medieval market scene - at the center, a fruit vendor's colorful stall, overflowing with fresh produce, apples glistening in the morning sun."

Impact: This starts with a broad setting (medieval market) and zooms into a specific element (fruit vendor’s stall), guiding the AI to create a detailed focal point within a defined context.

Do: Layer the details as you refine the prompt.

Effective Prompt: "An ancient library filled with shelves of old books, a golden chandelier above, and a large, world map spread across a central reading table."

Impact: This prompt provides a general setting (ancient library), then adds elements (shelves, chandelier), and ends with a specific detail (world map), creating a rich and immersive image.

Don'ts - Lack of Progressive Detailing
Don't: Jump into specifics without setting the scene.

Ineffective Prompt: "A crystal chandelier and a world map on a table, in a room."

Impact: This prompt lacks context, which may lead to a disjointed image where the items don’t seem to belong to a coherent space.

Don't: Be overly detailed from the start without establishing a setting.

Ineffective Prompt: "A golden chandelier with intricate filigree patterns and a world map with detailed topography."

Impact: This prompt dives into details without giving the AI information about the setting, possibly resulting in a detailed but contextually flat image.

"Spatial description" or "relative description" is crucial when crafting AI art prompts because such descriptions help in guiding the AI to understand and generate images with accurate positioning and relationships between elements. When you provide clear spatial relationships, the AI can better interpret how objects should appear in relation to one another, creating a coherent and visually logical scene.
Use specific descriptors for style and content.
Prompt: "A breathtaking landscape painting of the Scottish Highlands during sunset, with vibrant colors and a dramatic sky."

Impact: Generates a detailed image focusing on the quality of the light and the richness of the scene, in a painterly style.

Apply weight syntax to fine-tune details. Weight Syntax: In the context of AI art generation, 'weight syntax' refers to the use of numerical values alongside elements of your prompt to indicate their relative importance. This helps the AI prioritize certain aspects of the image. For instance, "(smiling:1.1)" suggests that the smile should be a prominent feature in the image.
Prompt: "A portrait of a young woman ((smiling:1.1)) with freckles."

Impact: The smile and freckles will be more pronounced due to the increased weight, making them focal points of the image.

Don't be vague or contradictory in your prompts.
Prompt: "A detailed photo of a cat, anime style."

Impact: Confusing the model by asking for a photo in anime style might result in a less coherent image, as photos are typically realistic and anime is a stylized art form.

Create a logical progression of details.
Prompt: "A serene spring morning in a Parisian cafe, with fresh croissants on the table, and the Eiffel Tower in the distant mist."

Impact: Offers a clear setting and progression, which helps the AI construct a scene with depth and relevant details.

Emphasize the mood or atmosphere.
Prompt: "An ethereal forest path, dappled with sunlight, evoking a sense of mystery and wonder."

Impact: Sets an emotional tone, guiding the AI to include elements that contribute to the intended mood.

Don't Mix too many styles or themes.
Negative Prompt: "A futuristic medieval castle with robots and knights, in a photorealistic manga style."

Impact: Combining conflicting themes and styles may result in a disjointed or cluttered image that lacks a clear focus.

Use culturally or historically accurate terms when needed.
Prompt: "A traditional Japanese tea ceremony, with participants wearing authentic kimonos."

Impact: Ensures the AI generates an image that respects the cultural or historical context of the scene.

Guide the AI on the focus of the image.
Prompt: "A close-up of a bee pollinating a vibrant sunflower, with a soft-focus background."

Impact: Directs the AI to focus on the bee and the sunflower, with a blurred background, creating a clear subject in the image.

Provide a concise, yet descriptive prompt that conveys the desired outcome without unnecessary verbosity. This means including essential details that define the subject, style, and mood of the image, while omitting extraneous information that does not contribute to the desired result. For instance, rather than simply saying 'a dog,' specify 'a golden retriever basking in the afternoon sun at a quiet beach,' which gives a clear image without being overly wordy.
Establish Visual Hierarchy: Visual hierarchy is essential in prompt crafting to direct the AI's attention to the most important elements of your image. By following these guidelines, your prompts will help the AI to create images with a clear focal point and a balanced composition. Use descriptive cues to dictate the prominence and relationship of objects:
Size: Indicate which elements should be large or small to suggest their importance.
Placement: Mention if something is in the foreground, middle ground, or background.
Contrast and Detail: Request more detail or higher contrast for important elements to make them stand out.
Example: "A towering lighthouse stands prominently in the foreground, its bright light contrasting against the dusk sky, while in the background, small ships dot the horizon."

Character Descriptions Without Naming: When describing characters, focus on their attributes, demeanor, and actions to convey who they are. Avoid using specific names that imply a pre-existing character the model wouldn't recognize. Describe characters by their traits, roles, or by a descriptive moniker that clearly communicates their essence or appearance. Example: Replace "Jacob, with a carefree and disheveled look," with "a carefree youth with disheveled hair." This way, you describe the character's key features without assuming the model's recognition of personal names.
Starting Your Prompt: Begin your prompt by directly setting the scene or introducing the action, without preambles such as "Envision a" or "depict a." The AI does not require such instructions to generate imagery. Start with a clear and engaging description of the environment, action, or subject matter you wish to see depicted.
Anatomy of a Prompt: The prompt structure should follow this structure: Subject, Detailed Imagery, Environment Description, Mood/Atmosphere Description, Style, Style Execution
Subject: The subject is the centerpiece of your image, demanding the viewer’s attention and defining the primary message. It could be:
•	Character: A person or creature, detailed with persona and context.
•	Object: Any inanimate item, grand or simple, with significance.
•	Scene: The larger environment setting the narrative stage.
•	Action: A dynamic occurrence infusing life into the image.
•	Emotion: The feeling or sentiment the image should invoke.
•	Position: Spatial arrangement of subjects within the scene.
Detailed Imagery: Adding Depth and Nuance. Enrich the subject with specific, engaging details such as:
•	Clothing: Describe attire with cultural or stylistic significance.
•	Expression: Convey emotions through facial and body language.
•	Color and Texture: Choose palettes and textures to set the mood.
•	Proportions and Perspective: Define the scale and viewpoint.
•	Interactions: Illustrate the relationship between different elements.
Environment Description: Setting the Stage. Craft the setting by detailing:
•	Indoor/Outdoor: Specify the primary environment.
•	Landscape: Describe geographical features or urban structures.
•	Weather and Time of Day: Set the scene with atmospheric conditions.
•	Background and Foreground: Add context and focus to the subject.
Mood/Atmosphere: The Soul of the Image. Evoke the intended emotional response by describing:
•	Emotion and Energy: The overall feeling or intensity of the scene.
•	Tension or Serenity: The dramatic or peaceful nature of the image.
Artistic Style: The Aesthetic Choice. Select your visual genre to set the stylistic tone, such as:
•	Anime to Photographic: Dictate the level of realism or stylization.
Style Execution: Bringing the Vision to Life. Detail the methods and tools for realizing the style, like:
•	Illustration Technique: Specify hand-drawn or digital methods.
•	Materials: Mention traditional or digital artistic tools.
Example of a “Subject, Detailed Imagery, Environment Description, Mood/Atmosphere Description, Style, Style Execution” Prompt Structure: “A bustling futuristic city with skyscrapers. Sleek, metallic surfaces with neon accents. Cars weaving through the cityscape. An electric atmosphere of innovation. Neon Punk aesthetic. Vibrant neon colors with sharp contrasts.”
Writing AI Art Prompts
// Whenever a description of an image is given.

// 1. Always mention the image type (photo, oil painting, watercolor painting, illustration, cartoon, drawing, vector, render, etc.) at the beginning of the caption. e.g. “a phot of a man eating an apple…” or “an Oil Painting of a dimly lit room”. Avoid more possibly ambiguous terms like “a photo capturing a man eating an apple…”

// 2. Make choices that may be insightful or unique sometimes.

// 3. Maintain the original prompt's intent and prioritize quality.

// The prompt must intricately describe every part of the image in concrete, objective detail. THINK about what the end goal of the description is, and extrapolate that to what would make satisfying images.

// All descriptions sent to me should be a paragraph of text that is extremely descriptive and detailed. Each should be more than 3 sentences long.

// If I request modifications to previous images, the captions should not simply be longer, but rather it should be refactored to integrate the suggestions into each of the captions.

// Clear Central Subject and actions: after stating the image type at the start of the prompt, clearly and concisely define the primary subject, actions, and location, so the focus is immediately established. You can add detail about each aspect of the image after the initial sentence with the clear subject.

// Detail Level and Structure: Start with an overarching description to provide context or set the scene. Proceed to describe specific elements or components of the image. Conclude with highlighting unique or symbolic features that provide deeper meaning to the artwork.

// Objective Description with Inferred Meaning: Use descriptions that are objective and avoid emotional or subjective terms.

// Avoid Ambiguity: Ensure descriptions are clear and avoid leaving major elements of the image to interpretation. Provide concrete details that give a strong sense of the artwork's visual components.

// Don’t be unnecessarily repetitive within a single caption.

// Don't reference things that aren't in the image like “as if the cameraman is taking the photo from atop a skyscraper.”

// Don't say things like "the main focus is" or "particular attention is given to…”. The structure of the prompt and the order in which things are described automatically imply what the image should focus on.

// Never mention "prompt" in front of the prompt

### END ###

Before responding and writing the perfect prompt for me, ask me 3 questions that will help you better understand the image I want to create to provide a even more accurate prompt. The 3 questions will be presented exactly like the template below. Nothing more than the content of the template until you get the response of the User, Wait for the user to respond to the questions or say GO before generating the prompt.

### RESPONSE TEMPLATE ###
I would like to ask you three questions to help me better understand your vision:

1. "Question 1"
2. "Question 2"
3. "Question 3"

If you don't want to add more information simply respond "GO". 
I will then generate the prompt for you.
When the response is generated click on the Image icon located at the bottom of the prompt to proceed with the image generation.
### END OF TEMPLATE ###

Remember, when you write the prompt, don't add Prompt or anything else that is not needed for the prompt in this response. No additional comment can be added in this response.

Below is the image description I want to acheive.

### IMAGE DESCRIPTION ###
[Describe your Image here]
### END ###