avatarGeeky Animals

Summary

The website content provides a quick update on Midjourney, detailing five key highlights, including the release of a free magazine edition, upcoming software versions, ongoing and future projects, and tips for prompt efficiency and image management.

Abstract

The latest Midjourney Weekly Office Hours reveal several significant updates for users. A free edition of Midjourney Magazine is available for a limited time with a special coupon code. The company is working on Midjourney V5.3, which may introduce aesthetic personality tests and higher resolution upscalers, while Midjourney V6 promises further enhancements in image quality and features. Ongoing projects include website updates, color customization using hex codes, and an image description initiative to train V6. The content also discusses the efficiency of prompts using Midjourney, with a focus on achieving desired outcomes with minimal GPU minutes, and provides guidance on managing image files using private servers. Additionally, it introduces methods for creating eye-catching featured images and encourages users to follow the author for future updates and tips.

Opinions

  • The author suggests that subscribing to Midjourney Magazine is beneficial, offering a free edition as an incentive.
  • The upcoming Midjourney V5.3 and V6 are anticipated to bring notable improvements, indicating the company's commitment to advancing its product.
  • The author emphasizes the importance of prompt efficiency and provides a methodology for measuring it, highlighting the value of minimizing GPU usage.
  • The use of keywords like "flat lay" and "knolling" is recommended for generating organized digital asset collections.
  • The author values user engagement and encourages readers to support the content by following on Medium and subscribing to email alerts.

Quick Midjourney Update: 5 key highlights to know (27th September 2023)

Keeping up with the rapid changes and developments

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

This image was created by employing a novel method for color-matching the subject and background, as described in Use Case 6 of the story “The Complete Guide to Midjourney Inpainting: From Basics to Advanced.” Vary (Region) A futuristic cyborg, fashion photography, glamorous — ar 3:2. (All images were created with Midjourney by the author.)

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

(1) Free edition of Midjourney Magazine

  • If you haven’t already subscribed to the magazine, use the coupon code O5ZQN1KP for a free edition (the code is limited to a certain number of issues and expires on October 5th).
  • Website for magazine subscriptions: https://mag.midjourney.com/
  • The new magazine (Issue 7) is shipping to magazine subscribers.
To receive a complimentary issue of Midjourney Magazine, enter the coupon code O5ZQN1KP.

(2) Midjourney V5.3

  • V5.3 might be released in October.
  • It will include an aesthetic personality test to help with aesthetic customization for image creation.
  • It could have an upscaler for higher resolution. The resolution could be 2K x 2K, and the widescreen version could be 4K. A version of upscaler with only minor changes to the original image is being considered.

(3) Midjourney V6

  • The release date is unknown. It could take 4 to 8 weeks, therefore unlikely in October.
  • The features include improved image quality, prompt understanding, better blends, higher resolutions, and potential text drawing capability.
  • A lengthy feature list may cause delays.

(4) Ongoing and future projects

  • Website update: the same as the previous update. The main objective is to roll out a stable version before adding more features.
  • The color customization feature using hex codes is in development with an unknown release date.
  • The image description initiative will ask users to provide detailed descriptions of images to train V6. Users may be rewarded with “fast hours” in exchange for participating.
  • The 3D feature will likely be introduced before the video.

(5) Other updates

(5A) Flat Lay and Knolling

  • Do you want to put together a collection of digital assets or items? Try the keywords mentioned in a Wei Mao story: “flat lay” and “knolling.” I used the keywords to create some vegetables, which worked as expected with satisfactory results. What do you think? Give it a try.
  • Based on my testing (>4 repeated prompting), Knolling is more likely to generate labels/numbers/captions/descriptions next to the images than Flat Lay. If you do not want the text, use Flat Lay instead.
  • Both keywords automatically isolate the items/assets on a blank background, which is a helpful feature.
  • The prompts:

/imagine prompt: 10 vintage botanical illustration of vegetables in the style of knolling --ar 3:2

/imagine prompt: 10 vintage botanical illustration of vegetables in the style of flat lay --ar 3:2

An example of using Knolling. /imagine prompt: 10 vintage botanical illustration of vegetables in the style of knolling — ar 3:2
An example of using Flat Lay. /imagine prompt: 10 vintage botanical illustration of vegetables in the style of flat lay — ar 3:2

(5B) Measuring Prompt Efficiency

  • Recently (14th Sep 2023), Clarinet published an article on measuring prompt efficiency to minimize the GPU minutes required to achieve desired results. The article was posted in the Prompt-faqs channel on Discord.
  • An efficient prompt will consume fewer GPU minutes to generate usable images.
  • The following steps are required to test a prompt’s efficiency: (1) Define the specific objective (a desirable outcome) or what you want to achieve. An image that fulfills this requirement means it is usable. (2) Roll the exact prompt four times to generate four image grids of 16 images. (3) Do not alter the parameters or /settings of the prompt. (4) Do not use the --seed parameter.
  • There are four types of prompt efficiency: (1) Robust (2) Reliable (3) Marginal (4) Inert
  • Robust — at least two or more images in every grid meet the objective. Your rolls consistently produce multiple usable images.
  • Reliable — Each grid contains at least one image that meets the objective. Your rolls always produce at least one usable image per roll.
  • Marginal — Only 1–2 of the sixteen images meet your requirements and are usable.
  • Inert — None of the images are usable.
  • Inert and marginal prompts are frequently the starting points for editing, enhancing, exploring, and experimenting to make them more reliable and robust.
  • Look for reasons why the prompts are unreliable — a spelling error? Incorrect punctuation mark placement? Did you accidentally copy-paste the wrong prompt?

Geeky Animals’ most recent stories:

You’re basically working in an operating system without folders if you don’t have a private server to generate and manage the images. Learn how to manage your image files on Discord using private servers (“folders”).

A simple method for creating eye-catching featured images with stock images and Midjourney. No prompting skill is required.

Previous Midjourney update on 13th September 2023.

Other stories

Latest news & updates on Midjourney:

Stories for Midjourney fans:

Simple and easy-to-follow guides for Midjourney Beginners

Stories about pets:

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

I hope you like this story!

Please give me a clap or leave a positive comment to let Medium’s algorithm know that this story is helpful and has contributed to its platform’s success.

Your support inspires me to create higher-quality content in the future!

Never miss a story. Follow me on Medium for updates, stories, and tips about Midjourney.

Sign up here for a free email alert when I post a new story.

Midjourney
Artificial Intelligence
Technology
Creativity
Design
Recommended from ReadMedium