MASSIVE MJ Updates + Video Course BETA

Fam,

I have to apologize…it’s been a while.

But for good reason.

We’ve got a lot to get to…so I won’t bore you with editorial.

Here’s what covering:

• Video Course Beta Launch.

• Consistent Characters in MJ.

• Pushing the CREF Function.

• Storyboarding.

• Stable Video.

• Midjourney’s Future Plans.

Yea, it’s a lot…but let’s do it.

Video Course BETA Launch

I’ve been working on this for 45 days straight.

(Explains my lack of posting here)

But it’s granular…it’s deep…and it covers both Discord + Alpha site tutorials.

Here’s more details:

• WTF is this: It’s a video course.

• Who is this for: Anyone who wants to get into the nitty gritty of Midjourney and hack image creation.

• When is the launch: Today.

• What Does it Include: 

  • (67) Video Tutorials,

  • (34) PDF Guides,

  • Monthly Content Updates,

  • (4) Cheatsheets,

  • 7-Hours worth of video content.

Your BETA access price is $147 USD. (This will eventually be $227)

The price is available for the first 50 people.

Pricing Tiers:

• $147 - First 50 purchasers

• $167 - Next 50 purchasers

• $227 - Full Launch.

It’s 99% complete…and I would more than appreciate any feedback.

Now that that’s out of the way…let’s get into it.

Consistent Character Update: Character Reference (CREF).

This Update was HUGE.

We’ve been yearning for it…and MJ gave it to us.

Here’s how to use it:

01. Create a base character in MJ
02. Copy image url
03. Create new prompt
04. Add [--cref] parameter
05. Paste image URL
06. Generate images.
07. Iterate prompts.

The base character is the most important. You don’t want to muddy the waters with too many additional elements. Keep your character the star of the show.

CREF Quick Hitters:

1. Use this on characters made from Midjourney images.

2. Not designed for real photos of people

3. Works similarly to image prompts 

4. Except it 'focuses' on the character traits

5. Precision is limited, 

6. It won't copy exact dimples / freckles / tattoos/ etc

7. Works on regular and Niji models.

Now with that being said…I’ve been having a lot success using a “character sheet” as the base character.

It works because it gives MJ multiple perspectives of the same character.

So more data = more direct results.

Here’s a quick prompt structure for Character sheets:

Prompt: character sheet, [subject], [wearing], studio setting, flat background, [camera/lens]; image split into 4 different pictures, shot from multiple angles --ar 16:9

Character Weight Parameter (--cw).

You can also balance your character reference with the Character weight parameter.

CW in Prompting:

• Use the --cw parameter to adjust importance

• Values: 0-100

• Default: 100

• cw 0: focuses on face (good for changing outfits / hair etc)

• cw 100: focuses the face, hair, and clothes

If you missed it…check out these two guides for the nitty gritty on CREF:

• Beginners Guide to Consistent Characters.

• Control Your Consistent Characters in Midjourney.

CREF with Objects:

This isn’t just limited to people…you can reference objects as well.

Typically, this has been working with “common objects” - Like cars, furniture, etc. The more intricate the object, the less likely it is to translate accordingly.

What’s been working best is creating the base “character” as a PNG or something with a transparent background.

Less noise in the background = more focused reference.

If you missed it…check this bad boy out: FULL POST.

It works very well with cars (specifically) and the subtle that are usually tough to translate from image to image stick. (which is nice).

It kept consistency on: Color, wheel spokes, break calipers, the grill, headlights, Etc

Here’s the Process:

1. Find/create source Image.

2. Use Image Prompt + Image Weight.

3. Utilize CREF for consistency.

4. Iterate shot type

Taking This Further…

You can do this with multiple objects…same process:

Storyboarding:

With CREF this became easier overnight….however it’s not as easy as it looks.

It’s complex and requires some “prompt competency” (I just made that term up)

But with the SORA coming, and the emergence of tools like Runway, Stable Video, Pika, etc…if you want to create ai video…you need a lot of good images.

If you want to build storyboards you’ll need to prompt for…

• Multiple character actions, settings, movements, etc.

So CREF will only take you so far…you need to text prompt correctly.

Here’s a few tips:

1. Start small and do one scene at a time

2. Try to make each scene (3-4) shots

Use a simple frame work for the entire story + each scene.

Framework:

• Who: [subject]

• What: [what they’re doing]

• Where: [setting]

• Start Point: [story opening]

• End Point: [story closing]

We cover this super in-depth in the VIDEO COURSE.

Once you have this down…

Teaming CREF + SREF is pure insanity…and Midjourney is pushing towards you being able to create entire worlds.

I have a weird brain…don’t knock me for it.

Stable Video:

Stable video was just released to the masses.

It’s a nice alternative.

Tested it and was pretty impressed - The object permanence is slightly better than runway and the objects remain sharp…still has a ways to go, but it’s worth checking out. (You get 100 free credits per day)

Here’s the link: https://stability.ai/stable-video

Midjourney’s Future Plans:

What the hell is Midjourney cooking up…

You can always get a sense of what’s coming next from their community polls…and it looks like they’re working on 3D capabilities + hardware.

You’re going to be able to build worlds and then experience them…we’re thinking small.

Here are the updates:

• Website

  • Working on the next rollout

  • Access to more people

• v7

  • Moving along (can’t believe i’m saying that)

  • Focusing on better image quality, prompt understanding, aesthetics

• Personalization

  • Trying to push model personalization out in 1-2 weeks

  • You will be able to turn the personalization on and off

• v6

  • May have one more update to improve body, hands, etc.

  • Looking at doing a style random

• Long Term

  • 50/50 chance 3D will come before video.

  • MJ video will be released before end of the year.

Have fun.

If you need help…DM me on Linkedin.

OR respond to this email with “HELP”…and we’ll chat.

Until next time…

RF