Summary

The Midjourney update from November 15, 2023, includes the release of the Niji·journey Style Tuner, the decision to skip version 5.3 to focus on version 6, and enhancements to the Midjourney website and image creation features.

Abstract

Midjourney has introduced several key updates in its latest release, with the highlight being the Niji·journey Style Tuner, which offers users a wide array of new style codes to experiment with. The team has decided to bypass the V5.3 update to expedite the development and release of the more advanced V6, which is expected to include aesthetic improvements and new features such as the ability to draw text aesthetically. The Midjourney website is set to receive improvements to its Explore page, filter features, and image download and collection capabilities, with a focus on catering to both new and experienced users. The update also addresses the community's feedback by planning to balance the needs of users who have generated over 10,000 images with those who are new to the platform. Additionally, the update discusses the potential for combining Midjourney with other systems like Adobe Firefly to leverage their collective strengths in image generation.

Opinions

The author suggests that the abundance of style codes from the Niji·journey Style Tuner, while potentially overwhelming, is a positive development, offering users a wealth of creative options.
David W Litwin, referenced in the content, demonstrates a positive view on using multiple image generation systems, such as Midjourney and Adobe Firefly, to enhance creative outcomes rather than seeing them as mutually exclusive.
The author expresses that the Midjourney V6 delay is justified, as it allows for fine-tuning and training on improved datasets to ensure a better user experience.
There is an acknowledgment that the majority of users (90%) have not created 10K images, indicating that website features should be developed with this majority in mind.
The author indicates a preference for the Midjourney community to embrace Discord, as it is a significant platform for user interaction and support, despite some new users' aversion to it.
The author encourages reader engagement by asking for claps and positive comments to signal to Medium's algorithm that the content is valuable and to inspire the creation of future high-quality content.

Quick Midjourney Update: 5 key highlights to know (15th November 2023)

Keeping up with the rapid changes and developments

Get a free email alert when a new story is published.

Style Tuner is constantly churning out new style codes. Users now have far too many style codes to experiment with. However, this is a good problem to have. /imagine prompt: a lady with a dog and a cat::1.5 desaturation, noisy, tracking lines, vintage looking 1980s VHS video still, 1980s, soft focused, 1980s VHS commercial — style 49KnRcwT1bMHxqTu03Qhij7tNmKGybNu5zuxnzX — ar 16:9

Here’s a quick rundown of the most recent Midjourney Weekly Office Hours:

(1) Upcoming releases

Niji·journey Style Tuner is now available. Join or message the bot here.
V5.3 of Midjourney will not be released. The priority now is to expedite the release of V6.
Hopefully, within the next two weeks, a website image download and collection feature will be available.
The website image creation feature will be available to power users who have generated over 10,000 images.
Midjourney now employs nearly 40 people, up from a team of 15 previously. More releases are expected in the coming months as manpower increases.

(2) Midjourney Website

The Explore page will be improved to allow a greater variety of images to be displayed.
The filter feature will be released, allowing you to filter images such as tile and square images.
According to reports, 90% of users did not create 10K images. The majority of users must be catered to when developing website features. Many newer users simply want to create images.
David estimates that 30–40% of new users dislike Discord. As a result, it is critical to ensure that new users (1,000 images generated) are familiar with the Discord interface first so that they can seek assistance from the community before creating images on the website.
Power users (>10K images) want a better image downloader, image collection, and organization features.
New website features must balance the needs of both new and experienced users.

(3) Midjourney V5.3

Midjourney V5.3 will not be released because the emphasis has shifted to accelerating the release of V6.
That means that the developer skips V5.3 and goes straight to V6.
Midjourney V5.3 is said to distract users due to its new prompting method.

(4) Midjourney V6 and V7

The release of Midjourney V6 has been postponed until early December, possibly before the holiday season. Or even wait until next year.
V6’s delay is due to:

More fine-tuning.
Training on new and improved datasets.
Checking on subtle details to ensure better understanding.

Midjourney V6 can aesthetically draw text.
The larger batch image generation feature (>4 images per grid) is now planned for V7 to allow for other more important priorities to be implemented in V6.
Following V6, the emphasis will be on style and character consistency. A skeletal posing feature is included.
There are more videos online than 3D images for training. Therefore, video development is faster than the 3D feature.
Most of the 3D images in the dataset were from video games, rendering the final output look like a video game rather than the standard Midjourney aesthetic.

(5) Other updates

(5.1) Accessing the Niji Style Tuner

Once you have joined the Niji·journey Discord server, you look for the Niji bot at the top right panel.

Click on the bot, and it will show you a button to “Add App.”

Click on the “Add App” to add the Niji·journey bot to your private server.

Pick your private server to add the Niji·journey bot.

If you are new to creating a private server, check out this story for more details and steps.

Once you have invited the Niji bot to your private server, type /tune to start creating Style Tuner in the Niji styles! (Use the /tune command from the green Niji icon)

Creating a Style Tuner page will consume your Fast Hours, even in /relax mode.

Here’s a “kitten” I created using the Niji Style Tuner.

You can also try out the Niji Tuner page I made.

Example of an image created using the style code from the Niji Style Tuner. /imagine prompt: kitten — niji — style k4Xpudaz — ar 16:9

(5.2) Midjourney list of Style Tuner pages

Since the introduction of Style Tuner, there has been a flood of artistic styles to choose from. Every day, new styles are discovered, and there are no words to describe them other than to assign them Style Codes.

Too many choices can be paralyzing. Try to pick the styles from the 128 Style Tuner page, and you’ll know what I meant.

But, hey, it’s a problem to have. There are so many things to try now. There are endless discoveries to be made when combining the style codes and tweaking the prompts.

What was the coolest thing that happened last weekend? Creating a Style Tuner page on the weekend was “free” with /relax mode. How long the developer will keep the /tune command working in /relax mode is unclear. I hope it lasts forever!

Midjourney has published a list of Tuner pages here. Check it out and see if you can find anything useful. I used a photography Tuner and a few style codes to create the featured image.

A photo created using a Style Tuner created by other Midjourney users. The style code is identical to the featured image. /imagine prompt: a lady with a dog::1.5 desaturation, noisy, tracking lines, vintage looking 1980s VHS video still, 1980s, soft focused, 1980s VHS commercial — style 49KnRcwT1bMHxqTu03Qhij7tNmKGybNu5zuxnzX — ar 16:9

(5.3) Using Midjourney + Adobe Firefly

Reading a story about using Midjourney and Adobe Firefly is refreshing.

Whenever another text-to-image system releases a new feature, the doomsayers say, “It is the end of Midjourney,” “Move over, Midjourney…” Hello, can’t we use more than one image generation system? Why does it have to be a one-and-only and mutually exclusive choice?

David W Litwin first used Adobe Firefly to create an image, which he then used as a reference and beautified with Midjourney. His process demonstrates how we can use the strengths of various image generation systems to create images for a project.

Here’s my take on the illustration using only Midjourney with direct prompting. /imagine prompt: a black-and-white illustration of a working lady using laptop, side view, sitting — ar 16:9

Geeky Animals’ most recent stories:

Tips for getting started in Midjourney.

7 things I’ll tell my best friend, who is just getting started with Midjourney

Tips for thriving in the fascinating world of Midjourney

bootcamp.uxdesign.cc

Previous Quick Midjourney Update and the style codes shared by Thaeyne.

Quick Midjourney Update: 5 key highlights to know (8th November 2023)

Keeping up with the rapid changes and developments

bootcamp.uxdesign.cc

Other stories

List of all Geeky The Cat’s Travel Log episodes:

Geeky The Cat’s Travel Log

Geeky, also known as Geeky The Cat, has exceptional intelligence and works at the Supreme Office of Agriculture. Geeky…

medium.com