Midjourney V6 is OUT! The Absolute Best AI Image Generator? | ALL NEW FEATURES

AI Catalyst
21 Dec 202305:26

TLDRMidjourney V6, the latest update to the AI image generator, has been released with significant improvements in photorealism and coherence. The update allows for more natural language prompts, eliminating the need for specific technical terms. It also introduces the ability to generate images with multiple characters and real people, although the latter may face future privacy restrictions. Midjourney V6 now supports adding short text to images and has enhanced its pixel art capabilities. While some features from V5 are temporarily unavailable, they are expected to be reintroduced. The new upscalers, 'creative' and 'subtle', offer different approaches to image enhancement. The website is currently a gallery for community images and will soon allow for image generation. For users, the $30 subscription may not yet be justified as the tool is still evolving, but the $10 subscription could suffice for casual use. Midjourney V6 is a step forward in AI image generation technology.

Takeaways

  • 🎉 Midjourney V6 has been released with improved photorealism and more coherent images.
  • 📈 The new version has significantly enhanced natural language understanding, allowing users to prompt as they would to a human.
  • 🖼 For photorealistic images, use the 'style raw' parameter and describe the subject simply.
  • 📝 Midjourney V6 can now generate multiple characters and images of real people, although this feature might be subject to privacy concerns.
  • 🔍 The ability to generate short text on images has been added, with promises of significant future improvements.
  • 🔧 V6 has improved pixel art generation and other image styles, but some users report it still ignores artist names and general art style descriptions.
  • 🚫 Some features from V5, like Zoom inpainting, 4X upscaling, and style tuning, are temporarily unavailable but will be reintroduced.
  • 🔄 Two new upscalers, 'creative' and 'subtle', are available, with the former adding details and the latter preserving the original image.
  • 🖥 Image prompts have been improved, allowing users to upload a reference image and generate new images based on it.
  • 🌐 The Midjourney website is being improved to serve as a gallery and allow image generation outside of Discord.
  • 💲 The $30 subscription is not yet considered worth it until all missing tools and features are fully restored and improved.

Q & A

  • What is the main focus of Midjourney V6 update?

    -The main focus of Midjourney V6 is to enhance photorealism, improve coherence and detail in images, and better understand natural language prompts.

  • How has Midjourney V6 improved its prompt understanding?

    -Midjourney V6 has improved its prompt understanding by allowing users to describe the image they want to create in a more natural language, without needing to use specific terms like '4K Unreal Engine' or camera model names.

  • What is the role of the 'style raw' parameter in generating photorealistic images?

    -The 'style raw' parameter is used to generate photorealistic images. By typing 'photorealistic' followed by the subject in the description, users can achieve a high-quality photorealistic image.

  • How does the 'stylized' parameter affect the image generation?

    -Lower values of the 'stylized' parameter improve prompt understanding, while higher values enhance the aesthetics of the generated images.

  • What is the new feature that allows Midjourney to generate multiple characters and images of real people?

    -Midjourney V6 has an enhanced ability to generate images featuring multiple characters and real people, such as actors, politicians, and celebrities. However, this feature may be subject to future changes due to privacy concerns.

  • What is the current state of the text generation feature on images?

    -The text generation feature on images is in its early stages and not very impressive at the moment, but the developers have promised significant improvements in the near future.

  • How has Midjourney V6 improved its pixel art image generation?

    -Midjourney V6 can now generate very high-quality and precise pixel art images, which was a weakness in the previous version.

  • What features from Midjourney V5 are currently unavailable in V6?

    -Features like Zoom inpainting, 4X upscaling, and style tuning are not available in V6 but are expected to be reintroduced gradually in the coming months.

  • What are the two new upscalers available in Midjourney V6?

    -The two new upscalers in Midjourney V6 are 'creative' and 'subtle'. The creative upscaler adds new details, while the subtle upscaler focuses on preserving the original image.

  • How has the image prompt process been improved in Midjourney V6?

    -Users can now upload a reference image, write a prompt, and generate new images based on the initial image, making the process more flexible and user-friendly.

  • What is the current status of the Midjourney website?

    -The Midjourney website is currently serving as a gallery where users can browse, rate, and examine community-created images and their prompts. It will soon allow image generation outside of Discord, but the full release date is not yet known.

  • Is the $30 subscription for Midjourney V6 worth it at the moment?

    -The $30 subscription may not be worth it until Midjourney V6 fully improves and regains all the missing tools and features. The $10 subscription is suggested for those who want to try it for fun.

Outlines

00:00

🚀 Mid Journey V6 Release and Features Overview

Mid Journey version 6 has been released, offering improved photorealism with more coherent and detailed images. The update enhances the software's understanding of natural language, allowing users to prompt it as they would a human, leading to closer results to their descriptions. The 'style raw' parameter is introduced for photorealistic images, and the 'stylized' parameter is used for better prompt understanding or aesthetics. The software now can generate multiple characters and images of real people, although this feature might face future restrictions due to privacy concerns. Additionally, it can generate short text on images, with improvements expected in the future. Compared to V5, V6 has made significant strides in creating pixel art and other image styles. However, some V5 features like zoom inpainting, 4X upscaling, and style tuning are temporarily unavailable but will be reintroduced. New upscalers, 'creative' and 'subtle', are available for image enhancement. The website is being improved as a gallery and image generation platform, with a beta version accessible to paid members. The video concludes by suggesting that the $30 subscription may not yet be worth it, advising potential users to wait for further improvements and tool reintroductions.

05:00

💡 Mid Journey V6 Subscription and Alternatives

The video script discusses whether the Mid Journey V6 subscription is worth the price. The narrator suggests that the $30 subscription may not yet be justified, recommending a wait-and-see approach until all the features and tools are fully improved and available. For those interested in experimenting with the software, the $10 subscription is deemed sufficient, although it comes with faster depletion of generation credits. The script also mentions DI 3 on Bing Image Creator as a free alternative that matches Mid Journey in all aspects except photorealism. The video promises a full guide on using Mid Journey V6 on their website and assures viewers that they will be updated as AI image generation technology evolves.

Mindmap

Keywords

💡Midjourney V6

Midjourney V6 refers to the latest version of the AI image generator, 'Midjourney'. It is the main subject of the video, where the host discusses its new features and improvements over previous versions. The video aims to assess whether the updated version is worth the subscription cost.

💡Photorealism

Photorealism in the context of the video refers to the quality of the generated images by Midjourney V6, which are very close to real-life photographs. The script mentions that the photorealism has been enhanced in this version, making the images more coherent and detailed.

💡Natural Language Understanding

Natural Language Understanding is the ability of Midjourney V6 to comprehend and interpret human language. The script highlights that the AI has significantly improved in this area, allowing users to describe the image they want to create in a more conversational manner, leading to more accurate results.

💡Prompt

A prompt is a description or request given to the AI to generate a specific image. The video explains that with Midjourney V6, users no longer need to use specific technical terms and can describe their desired image in a more natural way.

💡Style Raw Parameter

The Style Raw Parameter is a setting in Midjourney V6 that, when used, allows the AI to focus on generating photorealistic images. The video suggests using this parameter with a description of the subject to achieve high-quality photorealistic results.

💡Stylized Parameter

The Stylized Parameter is another setting in Midjourney V6 that affects the aesthetics of the generated images. Lower values are said to improve prompt understanding, while higher values enhance the visual appeal of the images.

💡Multiple Characters and Real People Images

This feature allows Midjourney V6 to generate images with multiple characters or real people, such as actors, politicians, and celebrities. The video mentions that this is an interesting update, but it also raises potential privacy concerns that might lead to its removal or censorship in the future.

💡Text on Images

The ability to generate short text on images is a new feature in Midjourney V6. Although the video suggests that this feature is not yet impressive, the developers have promised significant improvements in the near future.

💡Pixel Art Style

Pixel art is a digital art form where images are created on a pixel-by-pixel basis, often resulting in a blocky, low-resolution appearance. The video notes that Midjourney V6 has made significant strides in generating high-quality and precise pixel art images, which was a weakness in previous versions.

💡Upscaling

Upscaling in the context of Midjourney V6 refers to the process of increasing the resolution of an image while maintaining or enhancing its quality. The video introduces two new upscalers: 'Creative', which adds new details, and 'Subtle', which focuses on preserving the original image.

💡Image Prompts

Image prompts are reference images that users can upload to Midjourney V6 to guide the generation of new images. The video explains that this feature has been improved, allowing for more accurate and contextually relevant image creation based on the initial image provided.

💡Discord Polls

Discord Polls are a method used by the Midjourney community to vote on which features they would like to see prioritized in the development of the AI. The video mentions that several features from version 5 are not yet available in V6 but will be reintroduced based on community feedback.

Highlights

Midjourney V6 has been released with significant improvements in photorealism and coherence in generated images.

The new version has enhanced understanding of natural language, allowing users to describe images as they would to a human.

Midjourney V6 introduces the 'style raw' parameter for creating photorealistic images and the 'style' parameter for aesthetics.

The platform can now generate multiple characters and images of real people, although this feature might face privacy concerns in the future.

A new feature allows the generation of short text on images, with the promise of significant improvements in the near future.

Midjourney V6 can now create high-quality pixel art images, a significant improvement over previous versions.

Other image styles have also seen noticeable improvements in the latest version.

Users have reported that Midjourney V6 does not react to real artists' names or general art style descriptions, an issue that is expected to be addressed.

Several features from version 5, such as Zoom inpainting, 4X upscaling, and style tuning, are temporarily unavailable but will be reintroduced.

Two new upscalers, 'creative' and 'subtle', are available, each focusing on adding details or preserving the original image respectively.

Image prompts have been improved, allowing users to upload a reference image and generate new images based on it.

The Midjourney website is being improved to serve as a gallery and a platform for image generation outside of Discord.

Paid Midjourney members can access the beta version of the website for early testing.

The $30 subscription is not yet considered worth it due to missing tools and features, but the $10 subscription is suggested for fun.

Dolly 3 on Bing Image Creator is mentioned as a great free alternative to Midjourney, matching it in all aspects except photo realism.

A full guide on using Midjourney V6 will be available on the website, with updates as AI image generation technology evolves.