In this section, we will learn how to create images. We will learn how to effectively formulate requests for image generation and refine them. This process is similar to the operation of chatbots like ChatGPT. For example, by entering the word 'apple', a chatbot based on millions of diverse apple images creates the corresponding image.

Our main task for this section is to generate two images, the first one is the image of our hero, and the second one is the image of the background. To generate these images, we use LeonardoAI or MidJourney, depending on what you use. Personally, I use MidJourney - it's paid, $10 per month, but it does its job well, and there is a good free alternative - LeonardoAI.

How to write a good prompt: Define the main goal, the purpose of the image (what you want), then the background, then the environment (or what surrounds), and then details such as mood, color, etc.

You can read more about how to generate better images in the book - https://openart.ai/promptbook.

Effective prompts - we must precisely define and communicate to the bot what we want.

  1. Object - person, animal, character, location, item, etc.
  2. Type of background: photo, painting, illustration, sculpture, doodle, tapestry, etc.
  3. Environment: indoors, outdoors, on the moon, in Lviv, underwater, etc.
  4. Lighting: soft, natural, gloomy, neon, studio, etc.
  5. Color: bright, muted, bright, monochrome, colored, black and white, pastel, etc.
  6. Mood: calm, noisy, energetic, etc.
  7. Composition: Portrait, headshot, close-up, bird's-eye view, etc.

We use ChatGPT as a tool to generate prompts for MidJourney and other image generators.

Prompt for ChatGPT to generate our character: Describe the prompt for MidJourney for the portrait of an Ethiopian, an aristocrat of Kandakia, the Queen of Ethiopia

Then we use our generated prompt in ModJourney.

How to generate your photo in another style:

First, we describe the prompt in ChatGPT, then use it in MidJourney.

Prompt for ChatGPT: image.png Describe the man on the image

MidJourney prompt: <prompt from ChatGPT>, highly detailed emphasize all the details from the uploaded photo, especially the face

Wombo