There are a number of these MJ tutorial channels that are brilliant, and this tutorial is a good one in my personal assessment: no picture in picture redundancy, no ridiculous KZfaq 'humor', no long intros, clear articulation, polite language, and brevity! Five ⭐⭐⭐⭐⭐. Thanks. Will try this out.

Great video! I appreciate the detailed explanation on creating consistent characters using Mid-Journey versions 4 and 5. The different methods for obtaining the image seed in each version were really helpful. It's impressive how you demonstrated the character consistency in various environments, poses, and styles like cartoonish and 3D game engine looks. This technique will be very helpful for comic book artists and game developers who need their characters to remain consistent throughout their projects. Thanks for sharing, and congratulations on nearing the 600 subscriber milestone! Looking forward to more valuable content like this.

Something to try with "full body" would be to change your AR to be taller. Your wide screen AR is what is causing the AI to do half body. If you changed just your AI and put (full body photo) in parentheses you might generate the full thing. Then you can out paint if you want the image back to a wide 16:9. Just an idea.

These are *MUCH* clearer instructions than in some other videos. And it's short, sharp, and straight to the point, too! Like *ALL* videos should be!

Thank you for this straightforward, step-by-step video. Excellent useful information, and very well-presented. Thank you!

Bro this is the best vid for consistency I’ve seen! Bookmarked and much appreciation indeed!

This is the nearest I have got to consistency. Thank you.

Loved it. Amazing tutorial. Simple and to the point. God bless you mate. ❤

Straight to the point and very helpful! Good job!

Thanks! This is refreshingly clear.

This is very helpful. Thank you. You're good at knowing what to tell people to help them understand what to do and why to do it. Subscribed. Getting close to 1000!

I’ve been looking for this method for a while now. Thanks so much. 🙏🏼🙏🏼🙏🏼🌹👏🏼👏🏼👏🏼🔥🔥

This is insanely good! Thank you so much for this!

Tried this today... It really works. Thanks!

Love this! Veryyy inspiring & informative video! Thanks for the effort! :)

You explained so well in a very short span of time. Hope you hit million subscribers soon 👍

This is exactly what I was looking for. I didn’t know about the seed feature. Very useful, thanks for sharing.

Excellent & concise instructions

Awesome tips, love it! Thanks🤩

amazing fast, short, to the point video. Subbed!

Thank you. This was very helpful.

Wow. That last image of her, on the bottom right, is truly incredible.

I applaud you sir. Your explanation was concise and informative. I was not aware of the Seed tag. 🙏🏽

Thank you very much. V5 made me stressed because it never worked but with your method, it's working :)

Awesome vid!

It’s kinda working well. Thank you!

This is what I needed. Thanks a lot buddy. God bless you.

Great information!

Great tutorial! Thanks!

Been looking for this. ty :)

Thank you. This was easy to follow.

amazing tutorial, thx you Daz!

Thanks for your explanation, it's really clear and simple to understand.

Great tutorial! :) Thank you

Exactly what ave been looking for 👍

Great tutorial. thank you so much

Version 5.2 is able to give us seed number just by clicking Apps on the right above, then DM results Great video btw, keep it up buddy

Really good video! Thank you :)

Great video! Just subscribed 😊👌🏼

THANK YOU! Finally, a quick, here's what ya do, and this is what it does instead of an hour long video just to say far less, lol.

Very nice ! 🎉 I’ll try

Amazing!

I Feel like learned something very usefull. ❤ keep going bro

This is pretty consistent, but not quite there yet. To me still noticeable on first glance. Hopefully there will be an option to really use the exakt same face in the future.

Very cool thanks

I figured the best way to get a consistent character is creating different option sets based on how close the photo is supposed to be. So one for portrait, one for half portrait, one for full body. It's a bit of a hassle getting there, but it's worth it! I yet have to figure out how to add a second person to the consistent character without making them look same / similar. My character for example started wearing a dog's fur, when I wanted to create a little companion for her :D

Great job.

Good job.

It’s exactly what I am looking for❤❤

I'd send you a bottle of wine if I could, mate. I've needed this for a while now

👍🏻 great job

ohhhhhh thanks man finally i found the way for getting seed no in ver 5. thanks a lot

thank you so much for this! i just bought midjoureny V5 and i hope that it will make my anamorphic animal characters (like bugs bunny style) stay consistent for my children's story illustrations.

perfect tutorial

Great tutorial. Bummer that we have to do all of this and it still doesn't really get an exact match every time, just something kind of close. It's especially inconsistent with cartoon characters. Hopefully v6 will fix consistency.

Incredible video! Do you know how to do the same but with a real character with a photo I downloaded earlier?

Awesome. Saves to Promptbox

aaand... that's the end of modelling. We're doing well. Stunning chick.

Any advice on how to get consistent full heads? Mine always seem to get cut off and seldom show the full head, it's usually missing the top couple of inches; also how to get a full body shot, as again, I end up with waist up only, or similar. Thanks for this tutorial, it's extremely helpful.

thankyou for sharing! great video !! but what if i want to background to be the same... like i want to create a clothing lookbook, basiclly the background and the model should remain consistant while the clothing is changing as the prompts vary.

Portrait of Youth and Agawam Skyline both done with Midjourney 5😊

But with V.5 if you have to drop envelope on the grid for the seed, how do you specify which image on that grid? Or is that where the URL comes in? Thanks!

🎯 Key Takeaways for quick navigation: 00:00 🎨 Use Midjourney to create consistent characters for comic books, video games, or any project. 00:27 📝 Use tags and prompts to generate desired character variations in Midjourney. 01:24 🔗 In Midjourney 4, upscale the image and use the envelope icon to get the seed information for the character. In Midjourney 5, use the envelope reaction on the grid of four images. 02:46 ✍️ Write a formula in the prompt to instruct Midjourney to use a specific character with the desired seed. 04:01 ➡️ Change the character's actions, locations, and appearance by modifying the prompt and using different tags. 05:44 🙏 Thank you to viewers and subscribers for the support and engagement on the channel. Made with HARPA AI

Thanks for the great video, really insightful. I was wondering, how possible is it to create multiple characters and merge them in one scene.

I've seen lots of videos trying to make consistency of MJ character images, all failed to the "full body" prompt text, I guess we all just have to wait for the next release, to see if MJ fix this problem

Sooooooooooooooo Gooooooooooooooooooood

hey im trying this but its not keeping consistent.. similar yes but not like the same person any ideas?

Nice job! What was your cartoon prompt?

Is there a way to do this with just a downloaded photo? I made a character when Midjourney first came out, not thinking I would reuse it so I don't have the prompt or the url. I only have the photo.

Loving the video but for some reason I can't make this work with illustrations.. I get absolutely zero consistency..

This is where having the ability to use different systems work. You can take a few of those that are really really close and go into SD and use canny and pose along with the right parameters to get exactly what is needed with having truly similar characters with different poses. I don't really use just one system anymore - they all work hand in hand to get much better results than relying on one to get to the finish line.

I don't know if i can do it with leonardo AI but it's interesting to try.

hi. what I am not seeing is having the character doing an action. all I get is the character looking at the camera with a new background and clothes. anyway I can get the character to like stand at a bus stop looking down the street rather than looking at the camera?

Great tip for getting the seed! I wasted lots of time on that

ui, what a finding! Thank you so much with the help to be able again to get a seed of generated image in v5. Where did you find, that the envelope need to be send on the 4 images, instead of the upscaled image?

Hey does this work for niji 5 anime characters? And does the character have to be made in midjourney?

Hi! Everything i do, does not help me to bring my real photo face to the AI generated photo.. No even a remix mode. Any ideas?

does this still work? I don't see seed, job id or the envelop icon anywhere

Not able to get the seed from version 5. Any chance of being more specific?

Thank you! But I had an issue where the envelope command would not send me the job and seed info. I am using Discord in google chrome. I was able to get it to do that by: 1 : right clicking on the image 2 : scroll down to Apps 3 : choose DM results That had it display the job and seed info.

I haven't been able to acquire the seed for an image I generated. I've tried so many things! Any suggestions?

How to do this with a group of characters?

Thank you so much for this great method. Do you think there is a similar method with a picture you give to midjourney. For example : You provide a picture of you and ask midjourney to use your specific character in your specific picture ? All my attempts made midjourney create a character which looks like me but is not me. Thanks by advance

welp. curious how V6 will be with this.

4:45 "the full body tag didn't really work" me: oh, it did alright 😏

Thanks a ton for this; sadly, v5 just can't properly match specific characters even with this technique. I don't understand why they aren't providing the seed for the selected image when the 4 images can be so diverse/different. Sadly, I can't figure out how to create consistent characters with v5 in a clear fashion that's undeniably the same character. I think that's one of the biggest issues with Midjourney right now, and I sure hope v6 addresses the problem.

Anyway show how to create a video using mj v5. always wondered how people did it.

Upscale has no effect in version 5. You may save computing time in MJ when you not upscale and take your picture from the four composes.

Hi! On version 5, is the seed for all four images on the grid? how do you get it to select a certain image seed in the grid? thank you

If you've invited mid journey to your own server, you need to right click the server, privacy settings, enable direct messages

How would I get Midjourney to generate Me. I have thousands of pictures of me on my phone and it knows who I am - it tags me. How do I get that data to MJ. Im new.

Is your narration an AI generated version of your voice that you’ve trained? I’m curious as it’s consistently clear and occasionally the choice of words sounds more a written style than a spoken style. Obviously if it is really you then you talk very clearly! Well done! If not it’s still very impressive. It’s almost like the uncanny valley but for audio. Something just sounds slightly amiss for some reason. Maybe it’s just me 😂

Midjourney won't send the seed afters reacting with the envelope 😔😔😔

No matter what I do I can't seem to get this to work, has anyone tried this with melanated character prompts because most of the tutorials I watched seem to always use redheads.

how do you keep the exact outfit tho, say your trying to make an AI photoshoot , you want images to look like they're from the same place

This kind of stuff works only that well for realistic characters like yours. Try to do the same for Anime or Cartoon/Fantasy characters and you will get a complete different result. I struggle with that and it seems noting to work properly for now

Hi there, many thanks! Question, though. I have perfect character and seed, but she shares the render with another character. Is there a way to separate them?

Is there a way to get the seed information and prompt for previous pictures I created without having to dig through the discord?

No envelope symbol in my version! :-(

Is it not possible to issue seeds in Midjourney v5?😢😢😢😢If I request a seed issue and wait, the poop picture keeps appearing. Issue is not possible.