What's New in Midjourney October 2024: Image Editing and More

Midjourney released some exciting updates in October 2024. The biggest news is the new image editor. They also shared plans for personalization, faster image grids, and new model training. This post covers the key highlights from the latest announcements.

Edit Your Images with the New Midjourney Editor

A major new feature is Midjourney's image editor. For the first time, you can upload your own images and edit them directly within Midjourney. This opens up many creative possibilities.

Features of the New Editor

The new editor includes features you might know, like changing parts of an image (repainting) and adjusting the frame (reframing). It also adds a new feature called retexture. Retexture changes the style of your image while keeping the main shape and structure intact. This uses a control method to guide the transformation based on depth information.

Want to streamline your Midjourney workflow and explore advanced features like image editing more easily? Check out the Midjourney Automation Suite from TitanXT.

Who Can Use the Editor Now?

Midjourney is rolling out the editor carefully. It's currently available to three groups:

• Users who have created over 10,000 images total.

• Users with yearly subscriptions.

• Users who have subscribed continuously for the past 12 months.

This phased release helps them test a new moderation system.

Upcoming Changes: Personalization and Batch Size

More updates are coming soon to improve how you use Midjourney.

Easier Personalization

They are working on making it faster to set up your personalization profile. You'll also be able to create several different profiles for different styles or projects. A promising future update might let you upload a collection of images, like a mood board, and Midjourney could create a style profile based on that collection.

If you haven't explored personalization, head to the personalization page to get started. For a profile trained on anime styles, specifically for the Niji model, you can go to tasks and rank aesthetics for Niji.

Batch 8 Images

Batch 8 is planned, which will let you generate a grid of eight images instead of the current four. This will help you see more variations from a single prompt faster. It’s not certain yet if generating eight images will use more GPU time than four, but we will find out when it's released. Both the improved personalization and Batch 8 features are expected in the next few weeks.

Insights from Surveys and Model Progress

Midjourney is also analyzing data and improving its core models.

Survey Findings

The team is preparing to share insights from the demographic and psycho surveys you may have taken. They plan to release blog posts about what they learned. They might also create a tool that lets users look at how style preferences link to demographics.

V7 Model Updates

The V7 model has finished its first training phase and is now in the second phase. V7 has been trained on significantly more data than V6, about three times as much. This means V7 will understand more and produce better results. V7 is expected to be followed quickly by a V7.1 release, which will use an even larger dataset. There's hope to see V7 released before the end of the year.

Managing multiple model versions and features can be time-consuming. The Midjourney Automation Suite can help you manage your creative process efficiently.

Video and 3D Development

A V6 video model is still being developed. The focus for this model is image to video conversion, not text to video. Midjourney is stressing speed, simplicity, and good quality for this video release. There are no details yet on video length or resolution. The 3D development is also in training. It seems likely that the video model will be released before the 3D feature.

Handy Updates: Custom Shortcuts and Storytelling

Two other useful updates were mentioned.

Website Custom Shortcuts

If you’ve set up custom shortcuts using the prefer options commands on Discord, you can now use them on the Midjourney website. These shortcuts store things like character or style references and other prompt text. You still need to create the shortcuts on Discord, but being able to use them directly on the website makes prompting faster. For example, you could type --Wesley after the character reference parameter instead of typing out a long URL each time.

Storytelling Tool Preview

The team working on the storytelling tool is getting ready to share a preview. It sounds like this tool will focus more on creating entire worlds rather than just individual images. It's expected to have a different look and feel than the main Midjourney interface. Keep an eye on the events in the Midjourney Discord for updates on this.

Consider the TitanXT Midjourney Automation Suite to save time on repetitive tasks and manage your projects better across features like these.

Stay Updated with the Luma Events Calendar

There's a new Luma events calendar available. You can subscribe to this calendar to stay informed about all upcoming Midjourney events and announcements.

Conclusion

October brought significant updates to Midjourney, especially with the new image editor. With improvements to personalization, testing of Batch 8, progress on V7 and video models, and the ability to use custom shortcuts on the website, there's a lot to look forward to as Midjourney continues to develop.