Mastering the Art of AI Image Prompt Engineering
Written on
Imagine visiting an ice cream shop and selecting every flavor you love. It's thrilling, isn't it?
Similarly, AI image generators serve as those ice cream vendors, allowing you to manifest your creative visions with a variety of options.
Consider a prompt like “A lion and sheep together in a forest.” Such an image is impossible in the real world, yet AI art offers us limitless possibilities to visualize anything we can dream up.
> You are the creator, crafting your unique recipes.
Let’s explore how this works.
Be sure to follow this publication to stay updated with the latest in the realm of Generative AI, ensuring you never miss an important development.
GenAI For Everyone
Our objective is to make advanced AI technologies available to everyone through articles that simplify generative AI concepts.
Part 1: Basics
This article is a continuation from a previous one. For those who wish to start from there, please refer back.
All information presented here is derived from the platform PlaygroundAI.
Creating Stunning AI Images as You Envision
AI image generation can sometimes require patience. Grasping the fundamentals will empower you to excel in this field.
Part 2: Crafting Prompts
Engaging with AI image models is akin to interacting with young children; specificity and clarity are essential for them to produce the desired results, or it might lead to humorous outcomes.
If you instruct a child to “Take your seat,” they might interpret it quite literally.
Structure of a Prompt
Every prompt can be divided into three main components: > Style [Photographic, Artstyle-Watercolour, Abstract, 3D-model]
> Subject [What should be depicted in the image, including its action, mood, style, etc.]
> Additional Details [Cinemascope, film grain, highly detailed, 4k, vibrant]
Note that commas (,) are crucial for delineating your ideas within a prompt. This helps the AI model understand your request more effectively, resulting in a more fitting image.
What if the prompt isn’t structured properly? Just like children, AI models struggle to deliver precise outputs without clear guidance.
If you offer a vague prompt such as “A man smiling with happiness,” the AI may produce an image that diverges significantly from your expectations due to insufficient context.
Example
I need a high-quality photographic image of a little girl smiling and writing with a pen in a beautiful rainy forest.
Let’s approach this step-by-step for optimal results.
Step 1: A little girl writes in a book, smiling, amidst a rainy forest.
> Enhancing the prompt with style.
Step 2: Photographic, cinematic photo, a little girl writes in a book, smiling, amidst a rainy forest.
> Adding extra details to the prompt.
Step 3: Photographic, cinematic photo, a little girl writes in a book, smiling, amidst a rainy forest, 35mm photograph, film, bokeh, professional, 4k, highly detailed.
Check out the results from Steps 1, 2, and 3.
Applying Filters
Let’s apply some filters (available on the left panel of PlaygroundAI).
The filters utilized and the resulting images include (Macro realism, Copax Timeless V8, Pixel Art).
Generating Variations
Seeds are numerical sequences that instruct the AI on how to create an image. By altering any digit in the seed number, you can produce variations of the initial image.
For example: 235145777, 235145775, 235145776, 235145778, 235145779, 235145770.
Keeping Your Canvas Organized
In PlaygroundAI, you can easily generate various styles and maintain organization. This way, when you publish your article, retrieving images from the generated list becomes effortless.
Sample Prompts
Here are some prompts featured in the official PlaygroundAI YouTube tutorials that you can experiment with.
> Replace the placeholder text with {prompt} and utilize negative prompts to filter out unwanted details in the output.
#### Photographic Style: "cinematic photo {prompt}. 35mm photograph, film, bokeh, professional, 4k, highly detailed.”
Neg: "drawing, painting, crayon, sketch, graphite, impressionist, noisy, blurry, soft, deformed, ugly.”
#### Watercolour Style Style: "watercolor painting {prompt}. vibrant, beautiful, painterly, detailed, textural, artistic.”
Neg: "anime, photorealistic, 35mm film, deformed, glitch, low contrast, noisy.”
#### Cinematic Style: "cinematic film still {prompt}. shallow depth of field, vignette, highly detailed, high-budget Hollywood movie, bokeh, cinemascope, moody, epic, gorgeous, film grain, grainy.”
Neg: "anime, cartoon, graphic, text, painting, crayon, graphite, abstract, glitch, deformed, mutated, ugly, disfigured.”
#### 3-D Model Style: "professional 3D model {prompt}. octane render, highly detailed, volumetric, dramatic lighting.”
Neg: "ugly, deformed, noisy, low poly, blurry, painting.”
#### Abstract Style Style: "abstract style {prompt}. non-representational, colors and shapes, expression of feelings, imaginative, highly detailed.”
Neg: "realistic, photographic, figurative, concrete.”
Explore the complete list of prompts provided by PlaygroundAI for more ideas.
Stay tuned and subscribe for the latest updates.
If you enjoyed my article, please consider supporting me by clapping and sharing your thoughts.
Here are some additional articles you might find interesting:
- When Should You Believe You’ll Succeed as a Medium Writer?
- Want To Go Viral? Top-Writer Tips You Must Follow
- Have You Ran Out of Ideas to Write?
- Why Writing is Your Best Side Hustle?
- Two Articles a Day, 5-Day Accelerated Strategy in Medium
- Strategies to Boost Engagement on Your Medium Articles
- How I Covered a Day’s Rent With My Medium Earnings
- The One Golden Rule in Medium.
- Why You Must Believe in Something That You Haven’t Seen Yet?
- Lightening the Mind: A Simple Trick for Happiness
- I Can’t Do Anything: Break Free & Achieve Everything You Planned.