What's New in Midjourney? Updates on SREF, Video, and V8

Curious about the latest improvements coming to Midjourney? The team recently shared some big updates during their office hours. Get ready for enhancements to how you use references, create videos, and maybe even a brand new image model.

Enhancing How You Use References

Midjourney is making strides in how you work with image references. A new version of SREF and mood boards is on the horizon. The goal is for these to be very accurate and look great. Testing is happening soon, and they hope to release it quickly.

They are also improving the Omni Reference system. The team is training it with more data. A key focus is testing how to use several references at the same time when creating images. If both these improvements work well, they plan to combine the training results and release a major update.

Creating with multiple references can lead to amazing and unique results, but managing many attempts can be slow. Tools like the Midjourney Automation Suite from TitanXT can help here by handling repeated tasks and variations automatically, letting you focus on creative ideas.

[HH2]Progress on Midjourney Video[/H2]

Text-to-video generation is functioning internally, but their main focus is now on the image-to-video model. They are working on getting this model ready and aim for it to perform as well as, or better than, text-to-video.

It’s tough to say exactly when it will be ready. The internal plan is to finish training by the end of May and then move into testing. The team feels good about the progress. They think the videos will have motion and balance quality with cost. It's their first attempt at video, so they expect some things to improve in later versions.

Looking Towards Midjourney V8

Work is also happening on the next main image model, Version 8. This is expected to be a notable step up. It will use new hardware centers and a fresh model structure. The plan is likely to transition to V8 entirely rather than adding every new feature to V7.

Before V7 becomes the default model for everyone, they are trying to make it a bit faster. Optimization is vital to make sure there is enough computing power for users. They are also exploring even faster ways to generate images, quicker than the current draft mode.

Updates to the Interface

Midjourney is trying out a new interface aimed at professionals. It's designed for high-quality screens, like 4K monitors, offering a more polished appearance. They are also considering how to make the experience better for mobile users.

Managing complex prompts or generating a large number of images to find the perfect result can get complicated. Using an external tool like the Midjourney Automation Suite from TitanXT can simplify this process, letting you test many ideas quickly without manual effort.

Secret Projects

Finally, the team mentioned they are making progress on their secret projects. What these are remains a mystery for now, but it's exciting to know there's more innovation happening behind the scenes.

Summary

Many exciting developments are underway at Midjourney, including improved reference features, the push for an image-to-video model, and the next major image model, V8. These updates aim to make creating images and potentially video easier and better.

Staying up-to-date with Midjourney variations and tests can feel like a lot of work. Tools designed to help automate your workflow, such as the Midjourney Automation Suite, can help you manage multiple projects and experiments efficiently.