In today's digital world, text-to-image generation has become increasingly popular. Midjourney is an innovative app that, like OpenAI's DALLE-2 and Stable Diffusion's DreamStudio, generates images based on text prompts.

This article will explore advanced techniques for creating effective prompts, offering suggestions for where to place images and ideas for image content.

The Power of Descriptive Text Prompts

The more descriptive your text prompt, the more vibrant and unique the output image will be. For instance, using a raw and direct prompt like "a person wearing a hat and driving a car on the moon" would generate a specific and engaging image. Experimenting with different prompts and levels of detail can lead to a diverse range of outcomes.

‍

‍

‍

Advanced Options for Text Prompts

To create predictable and consistent image outputs, there are numerous advanced options available. This section will discuss various settings, such as:

1. Providing keywords for style: Using keywords associated with styles, artforms, designs, artists, or genres can influence the resulting image. For example, specifying an artist's name or a particular design style can create images reminiscent of that artist's work or design aesthetic.

Examples : "Anime style", "Pixar style", "Cyberpunk style", "Banksy style", "Picasso style", "Dali style", "Disney style"...

Use lighting styles like "studio lighting", "cinematic lighting", "hard light", 'long exposure", "glowing lights"....

‍

2. Stylizing output: By adjusting the "stylize" setting, you can control the degree of stylization in your generated image.

add to your prompt "--s <some number>"

‍

3. Using chaos for abstraction: The "chaos" setting allows you to increase or decrease the level of abstraction in the generated image, resulting in more or less abstract outcomes.

add to your prompt: "--chaos <some number>"

‍

4. Resolution: Specifying the output resolution using terms like "8K," "4K," "photorealistic," or "ultra detailed" can help achieve the desired image quality.

add to your prompt : --hd or --quality <some number>

‍

5. Aspect ratio: You can control the width and length ratio of the output image to generate images with specific dimensions or aspect ratios.

add your prompt : -- ar 4:3 or --ar 16:4 for example

or for custom image size, add "--w <some number> --h <some number>" (w = width, h = height)

‍

6. Filtering out words to remove unwanted elements from your art

add to your prompt : --no <element unwanted>

‍

Here are some prompts ideas and results from AI artists :
‍

‍

‍

‍

‍

‍

‍

‍

Using Images as Prompts

Midjourney also allows you to use images as prompts, either by passing a URL of the image or specifying weights for image prompts. By adjusting the weights, you can control the degree to which the generated image resembles the prompt image. This technique is useful when you want consistent outputs across multiple images.

‍

/imagine image url + prompt idea

/imagine image url 1 + image url 2

Conclusion:

Understanding and utilizing advanced techniques for crafting Midjourney prompts can lead to more consistent and creative image outputs. By experimenting with different text prompts, keywords, and settings, you can generate a wide range of stunning images that cater to your specific needs and preferences. The possibilities are endless, so have fun exploring the world of Midjourney's text-to-image generation!

Finally, to help you with you art generation, we would like to share with you a reference sheet containing Midjourney v5 photographers prompts and a github page with over 3000 keywords and styles for Midjourney.