Master Midjourney's Remix Feature and Create Amazing Images

In this newsletter, read about:

• 🕵️‍♀️ The Ultimate Guide to Editing Midjourney Images with Remix

• 🗞 News and Top Reads

• 📌 AI Art Tutorial: Create Text in Midjourney

• 🎨 Featured Artist: Jesse Payne

• 🖼 AI-Assisted Artwork of the Week

• 🤓 A Comprehensive Midjourney Guide

🕵️‍♀️ The Ultimate Guide to Editing Midjourney Images with Remix

Most Midjourney users don't realize how powerful the Remix feature is. It's a revolutionary tool for editing images with AI. You can use it to turn an illustration into a realistic photo, replace subjects, change the lighting effects, or guide your editing with image references – all while maintaining the overall composition and details of your initial image.

In this article, I'll show you an example of an image-generating workflow that leverages all of these opportunities.

But first, let’s check out /settings and make sure you have the Remix mode switched on.

Now we are ready to explore the full potential of Remix. Our workflow will be as follows:

1. Generating an illustration of a woman.

2. Changing it to a photo of a woman, while keeping the same composition and all the details.

3. Replacing a woman with a man.

4. Changing the lighting to Golden Hour.

5. Changing the environment using the reference image.

6. Final remixing for creating a coherent image.

Illustration of a Woman

We start with an illustration of a woman using the following prompt.

Looks nice, even though Midjourney kind of ignored the 1960s-style request.

Photo of a Woman

Now to change this illustration to a realistic photo, we choose Vary (Subtle) for the upscaled image, and in the pop-up Remix window, we replace “an illustration” with “a photo shoot on polaroid camera.”

Here is a realistic photo that retains the original image's exact composition and most of its details.

Photo of a Man

Next, we utilize the newly introduced Inpainting feature to substitute the woman with a man. To accomplish this, we perform the following steps:

• Select Vary (Region) on the upscaled image.

• Employ the Lasso tool to choose the woman.

• Substitute the original prompt with the following one.

Golden Hour Lighting

Now we again utilize the Vary (Subtle) button on the upscaled image to change the lighting effect. Here’s a prompt used for this Remix.

Even though we can see some indications of lighting change, the image doesn’t look great. Yet, it’s just a work in progress. Let’s continue.

Environment Change using the Reference Image

In this step, we combine the Inpainting feature with image referencing to replace our autumn tree with a green one.

First, we generate a reference image of the tree.

Next, we choose Vary (Region) on our previous image, select the tree area using both Rectangular and Lasso tools, and then, replace the prompt with the following one, where the link is the URL of our tree image.

Looks better, but still not good enough. Let’s have a final Remix.

Final Remix for Image Coherence

Often, Inpainting results in less consistent images in terms of style, lighting, and general mood. So, it’s usually helpful to do the final re-roll after Inpainting to let Midjourney adjust all the pieces together.

Here’s the prompt we use after choosing Vary (Subtle) on our latest image.

1960s style, a man, a green tree, a retro car, a city street, golden hour lighting – everything seems to be in place and consistently combined in one beautiful image.

Is it possible to achieve the same result by running this final prompt from scratch? Perhaps yes, but more likely, no. Nonetheless, the purpose of this guide is to showcase that Remix is not merely a tool for generating random re-rolls until acceptable results are obtained. It is, in fact, an excellent editing tool that you can leverage to gradually progress towards your perfect image.

🗞 News and Top Reads

• The US Copyright Office is seeking public input on AI and copyright.

  • The agency is opening the public comment period on August 30th with written comments due on October 18th and replies to be submitted to the Copyright Office by November 15th.

  • USCO is looking for input on issues like the use of copyrighted data in AI training, the copyrightability of AI-generated works, and the liability for copyright infringement involving AI.

• Google DeepMind is launching a beta version of SynthID, a tool for watermarking and identifying AI-generated images.

  • This technology embeds a digital watermark invisible to the human eye into images generated with Imagen, Google’s AI image generator.

  • It also can assess the likelihood of an image being created by Imagen.

• Microsoft wants to bring generative AI capabilities to its various Windows apps.

  • The experiments include bringing OCR to apps like Snipping Tool, Camera, and Photos.

  • The company is also playing with the idea of bringing generative AI to the Paint app.

📌 AI Art Tutorial: Create Text in Midjourney

In this guide, Nolan from Future Tech Pilot will walk you through the process of generating crystal-clear text using the Midjourney Inpainting feature.

🎨 Featured Artist: Jesse Payne

Jesse Payne has an extensive background in painting and drawing, having obtained a Bachelor of Fine Arts in Painting/Drawing from Indiana State University and a Master of Fine Arts in Painting from Northern Illinois University. Currently, he is an Associate Professor, leading the Drawing Curriculum in the Art Foundations Department at Virginia Commonwealth University, Doha, Qatar. AI art is just one of the art forms Jesse works with. In his Meet the Masters series, he merges the captivating allure of art history with cutting-edge AI technology to resurrect past artists and their lost works.

🖼 AI-Assisted Artwork of the Week

🤓 A Comprehensive Midjourney Guide

Share Kiki and Mozart

If you enjoy this newsletter and know someone who might also appreciate it, please feel free to share it with them. Let's spread the word about AI art and introduce more people to this fascinating field!