One of the best Midjourney V6 tutorials I've seen so far. I 'purchased' your Midjourney V6 Style Guide just this morning and worked through most of your examples and instructions within 2hrs. Then, I created a large spreadsheet in Mac Numbers (Excel on Windows would do the same) with the style categories and subcategories making the columns. For the cells in a columns, I then created pop down menus for the most common values, such as photography style, lighting, camera brand, etc. The last step was to create a new row that automatically creates on big prompt of all the individual cells. That makes creating new and changing existing prompts a breeze, because you do not need to fiddle around with a long text prompt in Midjourney, seeking for the field you want to change. Instead, I make the changes in the spreadsheet, and then copy the automatically updated prompt of the concatenated text values, copy it over to Midjourney's /imagine prompt - and voila! Just thinking whether I could even add the /imagine into my spreadsheet. With this approach you can beat everyone in quality and speed.
@@cyberjungle I tried it: yes, you can add the string "/imagine: " to your text prompt in Mac Numbers and copy the whole string to Discord - and discord takes it right away - wow!
If only now they add consistent characters with clothes and better posing you could literally create webtoons, manga's, and comic books with midjourney! I can't wait!
Learn color theory, composition, and lighting and you will be much better off for it. The people who will be successful in the future are storytellers with the fundamentals of cinematography and art. Aren’t going anywhere. Even with AI you still have to know what makes a good and bad shot for the human eye.
@@Jacksonmontyart Totally agree! As long as there is no AGI we as humans will have to be the ones deciding what the input is and validating the output. Be it creative or corporate jobs, it will take some time before AI will truly be self-sufficient without needing our help.
You could always learn how to use a drawing/painting app, like Procreate, Sketchbook, etc, and edit your images the old fashioned way. After all, an AI can't read your mind, and I doubt most people want to become writers, just to generate prompts that tell those AIs exactly what they really want. In the end, AI generated art is just another tool... One that can benefit an artist, but not really do a helluva lot for those with no skill, talent, or imagination.
Been watching a few of your videos today, really excellent quality and your understanding of film production and images shines through. The infographics and your example layouts are so useful, especially seeing inputs/outputs and comparison of value setting outputs and the central "feature comparison" graphic.
Very helpful CJ, thank you very much. And the point near the end, about weight of the stylise parameter: --sw {0 - 1000} is SO important. Crucial stuff! Ciao
Incredible!!! Thank you so much... Do you have any complete course to Midjourney that I could buy? I already purchased all the smaller ones but I would love to learn the advanced techniques from you, you are a wonderful teacher
great video! what about adding the reference image as an image prompt and then add the same image as a reference at the end? I mean... image URL , prompt, --sref image URL ... the end result will be closer to the image prompt or style reference?
TY. Great tutorial. How would you define the difference between adding a prompt weight to the end of a --sref ::[#] compared to adding--sw [0-1000] Assuming they both weight the image above the prompt, what (if anything), is the difference?
Please don't put your prompts in the top or bottom of your videos. When pausing the video in a browser and trying the parse them, the title covers the area at the top and the seek bar and controls covers the lower part making your prompt unreadable when the video is paused. Great content! Thanks!
10:30 there is too much anatomical detail in the area of the clavicle and neck muscles imv. As a newbie, how would you address that- is it a function of an original reference image?
@@cyberjungle Ah thanks so much for this. My nneurodivergent brain was thinking 'logically they should both amplify each other' and yet my results were not great. So this helps alot thanks.
For sref, I'm doing the exact same thing but I got an error said invalid parameter and Style reference must be an image URL, not text. Anyone knows how to resolve it? Thanks
@@cyberjungle Oh hey, thanks for answering! But like said I did the same prompt as you, maybe the image had to be MJ generated since URL must end in .png or equivalents, I uploaded the local file and it's not working
Do you need to adjust any settings for this to work? I am copying exactly how you do it, but my midjourney bot gives me an internal error. My URL links are also super long so idk if thats messing it up. I can reference 1 image but I cannot reference 2 images together
Thank you for all of this work. Is it possible to download your style guide into my version of Notion? Supplying the link is nice, but being able to import it into my instance of Notion would be evern better! Thanks so much!
100 people can write a text prompt and every one of them will be different. I find sref to be pretty useless right now. especially that it's still in alpha. But it is a step to consistency.
a nice tool but wish it was more open sourced. the pricing model it kinda pricy and the fact the webui only works for users who already generated a ton of images already instead of open to more people kinda sucks. a nice tool i suppose but for me not for my needs
Does MJ 6 even think of doing artistic styles anymore (more illustrative or painterly) as most of what I get despite trying to tell it more artistic and NOT NIJI (Anime) look, looks just like boring photos or movie stills. I have to go back to 5.2 just to get what I like. It's annoying. I pay per month too much it appears now. Not impressed with MJ6 or even the Update.
Completely DOES NOT work for historical images, ALL AI Image and Video creaters SUCK at accurate military History! Hell Midjourney threatened to throw me off if I continued to complain about it!!!!!!!!!!
I'm just wondering is it all worth it to learn how Midjourney works and to practice those prompts to reflect my expectations according to fact it's going to get more and more easy and user friendly to generate images with ultimate one-click-button solution haha
Hi Junglitos, here is the link to the updated Notion Library with Consistent Style section:
cihanunur.gumroad.com/l/midjourneyV6-style-guide
👋
Wow, this is a great idea... will check it out
@@LouisGedo 👋🏼
Thank you, thank you, thank you. 🙂
And I love that you've moved it to Notion. :-)
One of the best Midjourney V6 tutorials I've seen so far. I 'purchased' your Midjourney V6 Style Guide just this morning and worked through most of your examples and instructions within 2hrs.
Then, I created a large spreadsheet in Mac Numbers (Excel on Windows would do the same) with the style categories and subcategories making the columns. For the cells in a columns, I then created pop down menus for the most common values, such as photography style, lighting, camera brand, etc. The last step was to create a new row that automatically creates on big prompt of all the individual cells. That makes creating new and changing existing prompts a breeze, because you do not need to fiddle around with a long text prompt in Midjourney, seeking for the field you want to change. Instead, I make the changes in the spreadsheet, and then copy the automatically updated prompt of the concatenated text values, copy it over to Midjourney's /imagine prompt - and voila! Just thinking whether I could even add the /imagine into my spreadsheet. With this approach you can beat everyone in quality and speed.
Great idea and implementation
@@cyberjungle I tried it: yes, you can add the string "/imagine: " to your text prompt in Mac Numbers and copy the whole string to Discord - and discord takes it right away - wow!
If only now they add consistent characters with clothes and better posing you could literally create webtoons, manga's, and comic books with midjourney! I can't wait!
Learn color theory, composition, and lighting and you will be much better off for it. The people who will be successful in the future are storytellers with the fundamentals of cinematography and art. Aren’t going anywhere. Even with AI you still have to know what makes a good and bad shot for the human eye.
@@Jacksonmontyart Totally agree! As long as there is no AGI we as humans will have to be the ones deciding what the input is and validating the output. Be it creative or corporate jobs, it will take some time before AI will truly be self-sufficient without needing our help.
@@amykpop1 and even if it does get to that point, I can see human made art becoming much more valuable.
you can make that now
You could always learn how to use a drawing/painting app, like Procreate, Sketchbook, etc, and edit your images the old fashioned way.
After all, an AI can't read your mind, and I doubt most people want to become writers, just to generate prompts that tell those AIs exactly what they really want.
In the end, AI generated art is just another tool... One that can benefit an artist, but not really do a helluva lot for those with no skill, talent, or imagination.
Been watching a few of your videos today, really excellent quality and your understanding of film production and images shines through. The infographics and your example layouts are so useful, especially seeing inputs/outputs and comparison of value setting outputs and the central "feature comparison" graphic.
ngl watching this multiple times and taking notes lol. thank you for putting this together 🎉 keep up the great work it’s appreciated
Yes! finally what I've been waiting for. If only everyone knew the true value of this
I do
@@jackiewatkins1830what’s that
Immediate subscribe after 47 seconds: your combination of concept information graphics with examples is exactly what I've been looking for! Nice work.
Consistent styles are great. Now let’s hope we’ll get consistent characters soon.
Looking forward for it 👍🏼
Consistent Characters will change the game with consistent style 🥰💪🔥 please Midjourney give it to us
I needed to learn this. Thank you. Clear and extremely effective instruction.
Been waiting for a video like this! appreciate your awesome content :)
Stellar video, once again. Your work is greatly appreciated!
Very helpful CJ, thank you very much. And the point near the end, about weight of the stylise parameter: --sw {0 - 1000} is SO important. Crucial stuff! Ciao
Great video man! So much value, keep posting!!!!
Fantastic video and tips as always Ai bro
Your videos are incredibly useful and well made. Thanks for sharing your knowledge!
This is a crazy valuable tutorial wow thanks
10:57 the birds made me sub lol... also the really helpful information!
😂
You make the best MJ tutorials of all ⭐⭐⭐⭐⭐
Thank you so much!
Good job ! And thanks for your help
Thank you for the video and Midjourney V6 Style Guide!
Always helpful and well structured. Thank you!
Hi Cyberjungle, your Notion Library is so helpful.... thank you very much !!! 💪🤗
Great update - ) thank you for the video.
As always, a really good video. Thanks!
Great information ! Thank you !
Man this is awesome!
great help! thank you!!!
Incredibles! I need to learn to do This.
Amazing super clear! thanks
been waiting for product photography!!!’
Will be published in February
That’s Brilliant! Have a great week!
What a great review 💕
thanks for sharing bro
Thank you for this.
Incredible!!! Thank you so much... Do you have any complete course to Midjourney that I could buy? I already purchased all the smaller ones but I would love to learn the advanced techniques from you, you are a wonderful teacher
Thank you! I’m working on my Midjourney course 🙂
hey Cyber, I wish I found your channel earlier, Great Video & Super helpful tutorial.
Absolutely outstanding.
Fingers crossed this brings consistency to character details.
What about using multiple photos? Using the style from one, and reference from the other. What would that prompt look like?
if there is one midjourney chanel, this is the one !
THE ULTIMATE GAME CHANGER! COMIC BOOKS HERE I COME 🎉
very cool..trying it out now and it's amazing.
great video! what about adding the reference image as an image prompt and then add the same image as a reference at the end? I mean... image URL , prompt, --sref image URL ... the end result will be closer to the image prompt or style reference?
In this case, end result will be much closer to the Image Prompt 👍🏼
only the best content creator on midjourney
TY. Great tutorial. How would you define the difference between adding a prompt weight to the end of a --sref ::[#] compared to adding--sw [0-1000]
Assuming they both weight the image above the prompt, what (if anything), is the difference?
Please don't put your prompts in the top or bottom of your videos. When pausing the video in a browser and trying the parse them, the title covers the area at the top and the seek bar and controls covers the lower part making your prompt unreadable when the video is paused. Great content! Thanks!
Thanks for the feedback 👍🏼 I will be sure to fix that in the next video.
Like how you structure that video
10:30 there is too much anatomical detail in the area of the clavicle and neck muscles imv. As a newbie, how would you address that- is it a function of an original reference image?
Best place for style upgrades without losing quality
You really do a great job of explaining step-by-step. I happily subscribed and liked this video. Please keep up the good work!
great content! thx! :)
Can you both use image references as well as --sref? Or do this in turn perhaps? Or are they both better done separately?
I tested this. In this case, Image Reference bypasses sref. I recommend to use them separately 👍🏼
@@cyberjungle Ah thanks so much for this. My nneurodivergent brain was thinking 'logically they should both amplify each other' and yet my results were not great. So this helps alot thanks.
For sref, I'm doing the exact same thing but I got an error said invalid parameter and Style reference must be an image URL, not text. Anyone knows how to resolve it? Thanks
Hey be sure to copy paste image link after sref parameter and don’t forget to have a space in between
@@cyberjungle Oh hey, thanks for answering! But like said I did the same prompt as you, maybe the image had to be MJ generated since URL must end in .png or equivalents, I uploaded the local file and it's not working
Great vid! Thanks!
10:31 I would like to copy a pose from an image, and put my consistent character in that pose.
Same here. We need this feature! 👍🏼
I hope they fixed the eyebrows too, my wizards kept turning out weird.
Excellent!!!
Can we use cref and sref together in same prompt?
👏👏👏👏Amazing
Thank you!
Been glitching out too much lately, maybe now I can finish a DnD campaign backdrop without issues.
Thanks!!
Do you need to adjust any settings for this to work? I am copying exactly how you do it, but my midjourney bot gives me an internal error. My URL links are also super long so idk if thats messing it up. I can reference 1 image but I cannot reference 2 images together
Nice video
Thank you for all of this work. Is it possible to download your style guide into my version of Notion? Supplying the link is nice, but being able to import it into my instance of Notion would be evern better! Thanks so much!
Yes 🙌🏻 Now it’s possible
Thank you
Could you do it with stable diffution or COmfyUi
Amazing 😮
you are a genius! Can you make a your own GPT?
great content - thanks
can you have consistent style and character?
Yes, I will explain in the new video, dropping tomorrow
I LOVE THIS !!!T
I literally followed you becoz those bird were really threatening :'D
😂
Crazy
Think they’ll add video?
I’m certain they will add video but after yesterday’s OpenAI Sora announcement they have really challenging job to catch-up
100 people can write a text prompt and every one of them will be different. I find sref to be pretty useless right now. especially that it's still in alpha. But it is a step to consistency.
Can I use this on iPad?
Yes, they don’t have a native iPad app but you can use in the Discord App or on Safari
MJ is about to get 10x cooler.
a nice tool but wish it was more open sourced. the pricing model it kinda pricy and the fact the webui only works for users who already generated a ton of images already instead of open to more people kinda sucks. a nice tool i suppose but for me not for my needs
Does MJ 6 even think of doing artistic styles anymore (more illustrative or painterly) as most of what I get despite trying to tell it more artistic and NOT NIJI (Anime) look, looks just like boring photos or movie stills. I have to go back to 5.2 just to get what I like. It's annoying. I pay per month too much it appears now. Not impressed with MJ6 or even the Update.
Completely DOES NOT work for historical images, ALL AI Image and Video creaters SUCK at accurate military History! Hell Midjourney threatened to throw me off if I continued to complain about it!!!!!!!!!!
I think concept artists gonna lose a job in next year for sure
I think they will lose their job to human concept artists leveraging AI
Too complicated
I'm just wondering is it all worth it to learn how Midjourney works and to practice those prompts to reflect my expectations according to fact it's going to get more and more easy and user friendly to generate images with ultimate one-click-button solution haha
Thank you