Random Midjourney Update + Product Shots with Adobe.

Fam,

Another week…another Midjourney update.

I’ll keep the chit-chat to a minimum.

Here’s what we’re getting into:

1. Midjourney Mastery Course Update

2. MJ Update: Style Random

3. Juice Prompts w. Composition Tokens

4. Adobe’s New Play

5. Midjourney Office Hours

Let’s Boogie.

Midjourney Mastery Course: Last Shot (Extended):

I’m still making updates to the Midjourney Mastery Course…

So the BETA access has been extended.

I’ve added another (20) slots at the $167 price point.

The price will eventually be raised to $227.

Here’s the Details:

• BETA ACCESS INCLUDES:

  • (67) Video Tutorials

  • (14) Modules - Everything from signup-to-storyboarding

  • (36) PDF Guides

  • (4) Midjourney Cheat Sheets

  • Monthly Updates (New Videos + PDFs)

Your BETA Access Price is ($167) and is available for the next (20) people.

Ok…now that’s out of the way…let’s dive in.

MJ’s New Feature: Style Random

Under the cloak of darkness…Midjourney released a new feature.

It’s cool, but what you gain in creativity…you lose in control.

Let’s take a look:

Style Random:

What does it do?
+ It generates an image in a “random” style
+ Giving you unlimited aesthetic options

Here’s how to use it:
01. Create a prompt.
02. Add “--sref random” as a parameter
03. Hit Enter

Style Random Format:
vintage sports car --ar 3:2 --sref random

That’s it.

Thoughts:
+ It’s cool, but you have zero control
+ It’s a great way to burn fast hours lol
+ More fun than functional
+ There’s a ton of unique aesthetics
+ Less detailed prompt = More variation
+ You can transfer the style with SREF
+ Try with varying levels of stylize or chaos
+ Pair with weird parameter for some craziness

Prompt Example 1:
the perfect 40 m2 vintage shelby cobra, studio dark, checkered patterned floor --ar 2:1 --sref random --stylize 600

Juice Prompts with Composition Tokens:

One of the best kept secrets in Midjourney is using compositional elements to juice the hell out of your images.

They help construct unique visuals and they will definitely help you stand out.

If you missed it: Here’s the full post.

Composition:

Most people don’t take the time to learn composition…but…the compositional tokens in MJ are super powerful.

And just a few words can really stack the visuals nicely.

Here’s a few to try:

• Tonal Balance: Balancing the light/dark areas in an image.

• Color Blocking: Using bold areas of contrasting colors.

• Selective Color: Emphasizing color while the rest are b&w.

• Color Harmony: Combining colors that work well together.

• Tonal Transition: Colors transition and blend into one another

Prompt Example:

low angle shot, candid young boy in the Brazilian favelas, steps, doors, and windows, off-balanced composition, Canon EOS-1D X Mark II with 70-200mm lens, dappled sunlight, bold color blocking, extreme tonal balance between light and dark elements in the image --ar 2:1 --stylize 200

Quick Tip:

If the tokens are lacking firepower (within your prompt)…try explaining them in natural language.

For Example:

• Token: Tonal Balance

• Natural Language: extreme tonal balance between light and dark elements in the image

Also try using some “line” tokens - They definitely juice your images.

Line Tokens:

Lines are visual illusions in an image…they guide the viewer’s eye to the subject…here’s a few to try:

• Leading Lines: Guide attention to a specific part of the image.

• Horizontal Lines: Create a sense of tranquility.

• Vertical Lines: Imply strength, power, and elevation.

• Diagonal Lines: Suggest movement, and add drama.

• Implied Lines: A sequence of objects in a row.

• Repetitive Lines: Create rhythm or structure.

Prompt Example:

leading lines, low angle shot, Male DJ with green mohawk playing the turntable with a crowd, vibrant London {subway tube, underground club}, symmetrical composition, 70-200mm lens, vibrant tonal transition, stark contrasting light and dark elements in the image --ar 2:1 --stylize 200

Notice: How the lights, advertisements, the turntable knobs, and the wires lead directly to the DJ.

Here’s a link to DOWNLOAD FULL CAROUSEL + PROMPTS.

Adobe’s New Play

Adobe Firefly is now integrated into Photoshop Beta.

I’ve been a long-time user of Adobe’s products…and I’m a fan of how they’ve been competing in the Ai rat race.

It’s methodical…and they’re finding smart ways to integrate different features into different products

Check it out:

What it Means:

A ton of use cases - and Product Photography is something that can be achieved a little easier now.

You can remove and generate backgrounds easily and regenerate.

This isn’t “the solve” for Ai generated product photography…but…it’s lowered barrier to entry.

Product Photography:

Tried it…and did this in about 5 minutes:

Someone with real photoshop skills could go to town on these. If you need a quick social/website asset…this might be a good solution.

Generative Fill: Image Reference

You can also use an “image reference” within generative fill.

Again, I tested this very quickly…and looks like it still needs some work. I wouldn’t rely on this for anything professional…but you can see where it’s going.

If you’re an Adobe Creative Suite or FireFly user…here’s a good video that breaks down all the updates.

Midjourney Office Hours (4/24):

There’s some interesting nuggets here…let’s take a look:

Midjourney 6.1 Update

• Improvements to image quality, hands, coherence, and text rendering in progress

• Facing some technical hurdles but fixes are coming along

• Should be a really good 6.1 release once ready

Rooms: Opening Up Access

• Available to users with 1K images generated

• Will eventually add private rooms

Random Style Generation Codes

• Planning to release codes for duplicating random style generations more easily in the next few days

Personalized Models

• Working on shipping personalized models

• Currently working but inconsistent across users

• Trying to improve reliability and make it consistently awesome for everyone

Longer-Term Projects

• Teams working on V7, video models, and 3D models

• V7 and video need 6.1 improvements, so 6.1 is a stepping stone

• 3D held back by lack of data, but data collection efforts ramping up in the coming months

• 3D likely to come before video due to current video quality issues

• V7 being built to address video challenges in a better way

• Kicking off even more new projects and teams for long-term impact

12-Month Roadmap

• Focus on video, 3D, real-time capabilities

• Bringing them together for real-time 3D video and open-world non-interactive simulation

• Making the expensive capabilities cheaper over time

• Personalization is the major unknown in terms of its social effects and feedback loops

Have fun.

If you need help…DM me on Linkedin.

Until next time…

RF