Every Midjourney Feature Explained in 9 Minutes

Future Tech Pilot
16 May 202308:25

TLDRThe video provides a comprehensive guide to the features of Midjourney, a dynamic and frequently updated tool for image generation. Users can generate images in public channels and receive them privately via direct messages. The script covers various algorithms and styles, including version 4 with its four styles and the secret 'Cursed' style, version 5, and Niji Journey with its anime-based algorithm and additional styles. After image generation, users have options to re-roll, upscale, or create variations, and can make adjustments using the 'remix' feature. Parameters such as aspect ratio, chaos value, stylized value, quality, and seed number allow for customization. The script also explains advanced features like multi-prompts, negative prompts, and the ability to blend images. Users can save time with prompt shortcuts and suffixes, and manage their profile settings to control the visibility of their creations. The video concludes with tips on how to access help, FAQs, and user guides for further assistance.

Takeaways

  • 📬 React to a generated picture with an envelope emoji to receive a direct message for private generation.
  • 💡 Use the command `/imagine` in Discord to write your prompt, showcasing the capabilities of Midjourney with examples like 'cyborg, dinosaur'.
  • 🔍 Choose from various algorithms such as version 1 to 5, and explore styles like 4A, 4B, 4C, and a secret cursed style.
  • 🌈 With Niji Journey, you can select from different styles including expressive, cute, and scenic, which are based on an anime-trained algorithm.
  • ♻️ Use the re-roll feature to run the same prompt again, or upscale to generate a larger image.
  • 🔄 The variation button shows four slightly different images based on one of your pictures, and the remix button allows small adjustments.
  • 📏 Add parameters to your prompt like aspect ratio with `--ar`, chaos value with `-混沌` (which I assume is a placeholder for the correct command), and stylize value with `--stylize`.
  • 🎛️ Adjust the quality of the image with the `-q` parameter, which affects rendering time and cost.
  • 🌟 Use a seed number for consistency in image generation or the 'same seed' parameter for more similarity in a set of images.
  • ⏹️ The stop parameter allows you to halt the generation process at a specified percentage for different levels of image completion.
  • 📹 Include `--tile` in your prompt for seamless textures, and use multi-prompts with weights to emphasize certain elements.
  • 🔧 Utilize advanced parameters like `--test` for photorealism, `--test creative` for creativity, and `-HD` for high definition images.
  • ⚙️ Save time with prompt shortcuts using `--` followed by a chosen name, and suffixes that automatically append to all prompts.
  • 🔍 Use `--describe` to upload an image and generate prompts based on it, and copy job IDs to recreate or reference images.
  • 🎨 Blend up to five images together using the `/blend` command for composite creations.
  • 📊 Check your Midjourney profile for remaining fast time with `/info` and use `/relax` to switch to unlimited, albeit slower, relaxed generations.
  • ⛔ Pay for stealth mode to keep your generations private, as by default, they are public and accessible on the Midjourney website.
  • 📚 Access the FAQ and user guide through `/FAQ` and `/docs` for more information and assistance.

Q & A

  • How can I generate a picture privately using Midjourney?

    -You can react to a generated picture in a public channel with an envelope emoji, and Midjourney will send you a direct message to generate in private.

  • What is the command to write a prompt for generating an image in Midjourney?

    -The command to write a prompt in Midjourney is 'forward slash imagine', followed by the desired prompt text.

  • How many different algorithms are available for generating images in Midjourney?

    -There are multiple algorithms available, including versions one through five, each with its own sub-versions and styles.

  • What is the 'niji' version in Midjourney?

    -Niji is an algorithm in Midjourney that is trained on anime, offering different styles such as expressive, cute, and scenic.

  • What does the 're-roll' option do in Midjourney?

    -The 're-roll' option allows you to run the same prompt again to generate a new image.

  • How can I upscale an image generated with an older version of Midjourney?

    -When upscaling with an older version, you have access to different upscalers such as beta upscaler, light upscaler, and upscale to max.

  • What does the 'chaos' value in Midjourney determine?

    -The 'chaos' value determines the variety within the grid of generated images, with higher values leading to more diverse results.

  • What is the purpose of the 'stylize' parameter in Midjourney?

    -The 'stylize' parameter determines the creativity and artistic style of the image generation, with higher values resulting in more artistic but potentially less prompt-adherent images.

  • How can I adjust the aspect ratio of the generated images in Midjourney?

    -You can adjust the aspect ratio by adding '--AR' followed by a ratio of numbers, such as 1:2, 2:3, or 4:5.

  • What does the 'seed number' in Midjourney do?

    -The 'seed number' is a starting point for Midjourney to create your picture, and using the same seed number with a prompt will generate the same images.

  • How can I create a seamless texture using an image in Midjourney?

    -You can include '--tile' in your prompt to create seamless textures that can be used for patterns.

  • What is the 'multi-prompt' feature in Midjourney?

    -The 'multi-prompt' feature allows you to put a weight on certain parts of your prompt by using two colons, which lets Midjourney know to prioritize certain elements in the image generation.

Outlines

00:00

🚀 Mid-Journey Features and Generation Options

This paragraph introduces the various features of the Mid-Journey tool, which is frequently updated. It explains how to generate images in private channels using an envelope emoji, and the different algorithms available for image generation, such as versions 1 through 5, and the special 'cursed' style. It also covers the 'niji' version of the algorithm with its three additional styles. The paragraph details post-generation options like re-rolling, upscaling, and variations, as well as the 'remix' feature for making adjustments. Parameters that can be added to prompts are also discussed, including aspect ratio, chaos value, stylized value, quality parameter, seed number, and stop parameter. The paragraph concludes with additional features like creating videos of image generation, seamless textures, multi-prompts, and weighted prompts.

05:01

📚 Advanced Prompting and Mid-Journey Preferences

The second paragraph delves into more advanced features of Mid-Journey, including the 'repeat' parameter for running multiple attempts of a prompt and the use of squiggly brackets for permutations to save time. It discusses the ability to image prompt by combining an image with text, and the use of negative prompts to exclude certain elements from the generated images. The paragraph also covers additional parameters like 'test' for photorealism and 'HD' for high-definition images. It explains how to set preferences for prompt shortcuts and suffixes, use the 'describe' command to get image descriptions for prompts, copy job IDs for easy access, and blend multiple images together. The paragraph concludes with information on managing fast time, relaxing generations for free processing, accessing settings, and using the FAQ and documentation for further guidance.

Mindmap

Keywords

💡Mid-journey

Mid-journey refers to a feature within a software or application that allows users to generate images or content based on specific prompts. In the context of the video, it is a dynamic tool that is frequently updated and offers various algorithms for image generation, showcasing its evolution and capabilities.

💡Discord

Discord is a communication platform that supports text, voice, and video conversations. It is mentioned in the script as the medium through which the Mid-journey feature operates, allowing users to interact with the image generation tool by using specific commands and reactions.

💡Algorithm

An algorithm is a set of rules or procedures for solving problems or performing tasks. In the video, different versions of algorithms (version one to five) are discussed, each representing an advancement in the quality and style of the generated images, such as better-looking dinosaurs in version 4.

💡Niji Journey

Niji Journey is a specific feature or algorithm within Mid-journey that is trained on anime, resulting in image generation styles influenced by anime aesthetics. It is an example of how Mid-journey can adapt to different artistic styles.

💡Re-roll

Re-roll is an option within Mid-journey that allows users to run the same prompt again, potentially generating a new image. It is part of the customization process, offering users the chance to explore variations based on their initial input.

💡Upscale

Upscaling in the context of the video refers to the process of generating a larger version of an image. It is one of the options available after an image has been created, allowing for higher resolution outputs.

💡Variation

Variation is a feature that shows four slightly different images based on the original picture. It is a way to explore minor alterations and diversify the output from a single prompt.

💡Remix

Remix, when used in conjunction with the variation button, allows users to make small adjustments to their images. This feature enables fine-tuning of the generated content to better match the user's vision.

💡Parameters

Parameters are additional instructions or settings that can be added to a prompt to influence the outcome of the image generation. Examples include aspect ratio, chaos value, stylize value, and quality parameter, which affect various aspects of the generated image, such as its composition, diversity, artistic style, and rendering time.

💡Seed Number

A seed number is a numerical starting point used by Mid-journey to create an image. It ensures that if the same seed number is used with a prompt, the resulting image will be the same, allowing for reproducibility of specific image outcomes.

💡Stop Parameter

The stop parameter is used to halt the image generation process at a specified percentage, allowing users to control the level of detail and completion in the generated images. It can result in images that are barely started, blurry, or nearly finished, depending on the percentage chosen.

Highlights

Mid-journey is a feature that is frequently updated and allows for private image generation through Discord.

Reacting with an envelope emoji to a generated picture will result in a direct message with the image.

The command 'forward slash imagine' is used to start the image generation process with a custom prompt.

Different versions of algorithms are available, such as version one to five, each with unique styles.

Version 4 of the algorithm introduced four distinct styles including a 'cursed' style for humorous results.

Niji Journey is an algorithm trained on anime, offering styles like expressive, cute, and scenic.

After generation, options like re-roll, upscale, and variation allow for additional image manipulation.

The 'forward slash remix' command enables small adjustments to an existing image.

Older versions of mid-journey provide access to different upscalers like beta, light, and max.

The 'remaster' button applies an updated algorithm to enhance the image quality.

Parameters can be added to the prompt to adjust the aspect ratio, chaos value, stylization, and quality of the image.

A seed number can be specified for image generation to achieve consistent results.

The 'stop' parameter allows users to control the generation process by specifying a percentage stop point.

The 'dash dash video' parameter creates a video of the image generation process.

Seamless textures can be created with the 'dash dash tile' parameter for pattern designs.

Multi-prompts allow for weighted importance of different parts of the prompt for more control over the output.

The 'dash dash repeat' parameter enables multiple attempts of the same prompt for varied results.

Permutations can be created using squiggly brackets to save time and explore different prompt variations.

Image prompts can be combined with words for unique image-word combinations.

Negative prompts with 'dash dash no' allow for excluding specific elements from the generated images.

Additional parameters like 'dash dash test' and 'dash dash HD' offer photorealistic and high-definition options.

Custom prompt shortcuts can be created for frequently used phrases to save time.

Automatic additions to prompts can be set using 'forward slash prefer' for consistent output.

The 'forward slash describe' command allows mid-journey to generate prompts based on an uploaded image.

Old job IDs can be used to recreate and retrieve past images as if they were freshly generated.

The 'forward slash blend' command combines up to five images for a blended result.

Information on the user's mid-journey profile and remaining fast time can be viewed with 'forward slash info'.

Fast time can be replenished with the 'forward slash relax' command for slower but unlimited free generations.