Welcome to my very first newsletter. Every issue will explore new topics, aiming to give you a better understanding of AI and tech. So let's get started on this (Mid)journey together!
This month, I'm kicking things off with a closer look at Midjourney's advanced features. I'm here to share insights and tips that will help you get creative and make the most of this powerful tool.
How to prompt right with Midjourney
- Prompt length:?Short prompts yield images in Midjourney’s default style. Longer prompts offer uniqueness but don't overcomplicate. Focus on the main concepts.
- Grammar and word choice: Midjourney doesn't understand grammar and sentence structure like humans. Choose specific synonyms for better results and minimize word count for a stronger influence. Commas, brackets, and hyphens can organize thoughts but aren't always interpreted reliably.
- Describing your vision: Focus on what you want rather than what you don’t. For example, specifying “no cake” might still result in a cake in the image. If certain objects should be excluded, use advanced prompting with parameters like?--no.
- Details matter: Be specific about the subject, medium, environment, lighting, color, mood, and composition.
- Use of collective nouns: Specificity is key. Instead of general terms like "cats," use specific numbers or collective nouns like "three cats" or "a flock of birds"
/BLEND vs /DESCRIBE COMMANDS
What /blend does:
The?/blend?command in Midjourney allows you to quickly upload 2 to 5 images. It then analyzes the concepts and aesthetics of each image and merges them to create a new, unique image. This command is similar to using multiple image prompts with /imagine, but it is optimized for easier use on mobile devices.
How to use /blend:
- After entering the /blend command, you'll be prompted to upload two images.
- You can drag and drop images from your hard drive or add images from your photo library on a mobile device.
- To add more than two images, select the optional/options field and choose image3, image4, or image5.
- Note that the /blend command might take longer to start than others because the images need to be uploaded first.
- For the best results, upload images that are the same aspect ratio as your desired result.
Image Ratio Options:
- By default, blended images have a 1:1 aspect ratio.
- You can change this by using the dimensions field to select a square (1:1), portrait (2:3), or landscape (3:2) aspect ratio.
- Custom suffixes added to /blend prompts will work like any other /imagine prompt, and any aspect ratio specified in the /blend command will override aspect ratios in a custom suffix
What /describe does:
- Functionality: The /describe command lets you upload an image and generates four potential prompts based on that image. This feature is useful for exploring new vocabulary and aesthetic movements inspired by the uploaded image.
- Output: It produces prompts that are inspirational and suggestive. However, it's important to note that /describe cannot be used to recreate an uploaded image exactly. It's more about inspiration than replication.
- Additional Feature: The command also returns the aspect ratio for uploaded images, providing useful information about the image's dimensions.
??SPECIAL TRICKS
Add weights to your prompts
With the double colon(::), you can help Midjourney to interpret terms separately. This tip is handy when you want to increase or decrease the importance of certain terms in your prompt.
An example taken from the Midjoureney is the ‘space ship’ image. If the prompt is separated into two parts,?space:: ship, then each concept is considered separately, generating a ship sailing through space.
That's a wrap for today! In our theIntelligent course we delve deeper into AI strategies and tips. Make sure to jump onboard: https://theintelligent.ai/
