Finally! I've been looking for a video like this and I was having so much trouble understanding why my photo uploads were coming out so crappy. You explained everything so logically and in such great detail and your visuals made it so much easier to comprehend than many others I've watched. You're a NBT--Natural Born Teacher. Thank you!
EXCELLENT. Best tutorials ive seen. You show the action for every thing you describe. CLEAR and noit rushed. SILENCE between statements gives listener time to process new info. KEY. YOU actually know how to teach. GREAT job ! Keep on. SUBSCRIBED
I knew i could use a reference photo but i didn't know i could use multiple! Thanks alot, you just upped my game! Just tested it out with some of the old images i had made and the results where great.
Aa i read some of the comments it seems that a lot of artists are discouraged or even fearful of AI technology. instead of embracing or enhancing their craft they view AI Art as a threat. But this is life and life is about change and all those who do not learn to adapt get crushed and those that do flourish. Never give up your craft because of fear utilize your imagination to enhance your creativity. And stop being afraid of life embrace it and use this tool to expand your skills
This is what I've been looking for when it comes to leveraging images, I love it! I like the use the same image three times I would have never thought of that. So, Thank you! I will subscribe to your channel and share your content with the people who want to create NFTs.
Thank you! As a crypto enjoyooor, I'm planning on releasing a video on creating an entire NFT collection with AI. So that will be very relevant to them.
This was a tutorial and somehow it felt soothing to listen to it, not to mention that it was actually helpfull and that alot ☺️ Thanks alot, please making these, I'm a sub.
You could also use 3 different image links and add: a space, :: and a weight, so you can choose what links get the most influence so link1 ::1 link2 ::2 link3 ::1 L ink 2 will become dominant
These are great videos. Thanks again for your help with my hexagon plate on discord. I'm going to use the image that you created as image reference and play with the weights.
My most recent focus with a new AI tool called Blue Willow. It produced results that completely astounded me. If you could create a series tutorial like this for it, that would be fantastic for the beginners and it's completely free
Hey so I've been trying to get MidJourney to create some designs based off of a balisong but it never works. Even when I use an example picture, it disconfigures it and takes the pieces apart and moves them all around in a weird way every time I try it. Do you know of any way to fix this? Thanks in advance
Very insightful! Could you tell me the difference between increasing iw (1 to 2 to 5 to 10000) and increasing the number of reference links (I think they both make the results have more similar features of the original)
It is actually fascinating to me how the art community of our generation is basically "feeding" AI by collecting data + crypted informations. Somehow, we are currently developing the foundations that will allow ultra comprehensive conversations with "AI". How mesmerizing..
@@blahblahbear1429 Art became cheap and democratic. For better or worse. I have sufficient talent to be an artist myself but never pursued this venue cause there have been too many artists already. I enjoy AI generators immensely.
@@brexitgreens watch many artists in the future, including children, lose hope and passion to try art because “AI just does it better.” It’s really fucking sad.
@@blahblahbear1429 You're right. There are new AI programs able to interpret any style. The sad part is realizing devs are literally stealing data from notorious platforms such as artstation or pixiv.. Anyone can copy the style of notorious artists like Stanley Lau 'artgerm' or even artstyles fom the most famous renaissance painters. I'm geniunely ok with it because ten thousand years from now, none of us will be around anyway but AI will remain somehow so...probably for centuries and beyond, who knows
Is it possible to give the shape of the ref image a low weight. But the color a high weight? Because the clouds you made also changed shape based on the ref.
Midjourney now analyses the images. Uploading 3 times exactly the same image as reference is prohibited in the rules. It can get your access revoked. They have to differ from eachother.
thanks for your work, great job really. I've just tried the bot withe the 3 same images sequence, but Midjoiurney bot return this : One or more of your image prompts are the same image. Circumventing this filter to violate our rules may result in your access being revoked. -- is not normal?
Hello, My free trial has ended, and I'm unable to sign up for the paid plan since I don't have a credit card and Stripe isn't currently accessible in Kenya. I only have PayPal, Please advise on how to resolve this.
Hi Kaizen! Not sure I understand your question. But if you are referring to the fact that I have many prompts in the text box - that was purely for demonstration on the screen, I ran each individually.
so i'm getting errors with --iw is there a specific version of MidJourney that you are using for this video... I am trying to use --v 4 with the --iw prompt
Great tutorial. I am watching it 4 months later and there are some changes, or maybe I am doing something wrong. When I try to prompt 3 of the same picture, I get a message that filters prevent that prompt for multiple same pictures.
Great question. Yes you can do that by separating the images with :: and following that up with a number. Similar to text weights (have a tutorial on that too)
I tried this with 3 of the same image and I got this message using. Each was a unique link. Not sure what I didn wrong. 'Duplicate images detected One or more of your image prompts are the same image. Circumventing this filter to violate our rules may result in your access being revoked. Our image filter is currently a work in progress, and may have false positives.'
question: Can this work the same way for a full body picture of a character? Example: a full body reference pic of Green Lantern Kilowog from DC comics
It could, although it will be a lot more difficult - the more details you want Midjourney to get right the harder it will be. My example was a very basic composition, thus the results were very similar. Give it a try!
You can just drag and drop the image if you are on your desktop. If on mobile, open the channel you want to send it to and then click "+" sign next to your text box at the bottom.
I tried to use multiple images and it flagged that i have used multiple duplicate images and that if I try to circumvent this I may be banned access from Midjourney. Where did I go wrong? All the images had different links. Also before that it said i cant give any weight in v4. not sure what that means
"One or more of your image prompts are the same image. Circumventing this filter to violate our rules may result in your access being revoked. " I got this warning when provided the same images with different links. and also the --iw is not available in V4.
Interesting. I play around with pics of me for future album covers. And so far (3 weeks in) latest from today, following all suggestions, is, that I (a blond man middle long hair) created the black and white portrait of a dark haired woman. ::iw 5. That is... strange. You blond woman result is 100 x better.
Just tried the 3 images from of me from the same photo session, very similar angled shots, plus ::iw 1. The grid produces 1 Landscape and 3 abstract pictures. Humph. 🙂
Helpful video. Just what i was looking for to increase likeness of end result. But just received this message from MIDJOURNEY when i try to upload multiple same image like you suggested 👇👇 Duplicate images detected One or more of your image prompts are the same image. Circumventing this filter to violate our rules may result in your access being revoked. _______________ Suggestions? Thanks.
Using the same image multiple times as a reference doesn't seem allowed anymore. It says "Duplicate images detected One or more of your image prompts are the same image. Circumventing this filter to violate our rules may result in your access being revoked. Our image filter is currently a work in progress, and may have false positives."
Midjourney should begin making plans to raise their 25 free limit right away. As an alternative, AI technologies like Blue Willow will become more and more popular over time. Blue Willow, in my opinion, will be the one to take down MidJourney
Thanks for your comment Katie! It is included, I simply dragged and dropped the images from my Downloads folder to Discord and hit Enter (that was my downloads folder you saw on the top right)
Not sure if anyone else is getting this result now: Duplicate images detected One or more of your image prompts are the same image. Circumventing this filter to violate our rules may result in your access being revoked. Our image filter is currently a work in progress, and may have false positives.
I've spent hours and hours, dozens of photos, trying to get art based on it, and it simply won't work. It works just fine with photos found online though. Videos like this are misleading, and are the cause of me signing up and finding out it was absolutely pointless.
@@flickwtchr much better, especially when i have geeks who think theyre hot shit like you because they can write combination of words into a text bar LOL
![Are you writing your Text Prompts WRONG? Midjourney AI [Prompt Crafting Tutorial] - Ep.1](http://i.ytimg.com/vi/2QTXTp-_5JY/mqdefault.jpg)
![Are you writing your Text Prompts WRONG? Midjourney AI [Prompt Crafting Tutorial] - Ep.1](/img/tr.png)
Finally! I've been looking for a video like this and I was having so much trouble understanding why my photo uploads were coming out so crappy. You explained everything so logically and in such great detail and your visuals made it so much easier to comprehend than many others I've watched. You're a NBT--Natural Born Teacher. Thank you!
EXCELLENT. Best tutorials ive seen. You show the action for every thing you describe. CLEAR and noit rushed. SILENCE between statements gives listener time to process new info. KEY.
YOU actually know how to teach. GREAT job ! Keep on. SUBSCRIBED
Best video so far
I knew i could use a reference photo but i didn't know i could use multiple! Thanks alot, you just upped my game! Just tested it out with some of the old images i had made and the results where great.
Aa i read some of the comments it seems that a lot of artists are discouraged or even fearful of AI technology. instead of embracing or enhancing their craft they view AI Art as a threat. But this is life and life is about change and all those who do not learn to adapt get crushed and those that do flourish. Never give up your craft because of fear utilize your imagination to enhance your creativity. And stop being afraid of life embrace it and use this tool to expand your skills
This is what I've been looking for when it comes to leveraging images, I love it! I like the use the same image three times I would have never thought of that. So, Thank you! I will subscribe to your channel and share your content with the people who want to create NFTs.
Thank you! As a crypto enjoyooor, I'm planning on releasing a video on creating an entire NFT collection with AI. So that will be very relevant to them.
Dreambooth is way more competent for consistency. And it’s free.
Thank you a very good video explaining things
Thank you so much I did not knew I could use more than one image at a time! you blew my mind lol.
Thank you for watching! Glad it was helpful haha
This was a tutorial and somehow it felt soothing to listen to it, not to mention that it was actually helpfull and that alot ☺️
Thanks alot, please making these, I'm a sub.
This video is exactly what I was looking for, Thank you!
Thank you for this video! Exactly what I was looking for 😄
Thanks for watching!
So image weight cannot be used with v4, is it possible to use them with v3 then re-run as v4?
Very well done, thank you. Saw this in the midjourney discord
Thank you for watching and glad you found some value!
Well explained - straight to the point!
You could also use 3 different image links and add: a space, :: and a weight, so you can choose what links get the most influence so link1 ::1 link2 ::2 link3 ::1 L ink 2 will become dominant
You can indeed and its powerful BUT unfortunately you can't use :: with test and testp...
I didnt know Midjourney had an image reference function - excellent! I'm excited to try it out. Thank you for sharing!
Thank you for watching!
This was indeed quite helpful, thank you!!
Thanks for watching!
Excellent
Woww, thanks!!
cannot have image weights with version 4
These are great videos. Thanks again for your help with my hexagon plate on discord. I'm going to use the image that you created as image reference and play with the weights.
Glad I could help!
My most recent focus with a new AI tool called Blue Willow. It produced results that completely astounded me. If you could create a series tutorial like this for it, that would be fantastic for the beginners and it's completely free
Many thankse!
very good thanks
Great Video , straight to the point
But it's not working with midjourney V4
can U show us a way of how to use --iw with V4
Important lesson. Thanks, guru.
Hey so I've been trying to get MidJourney to create some designs based off of a balisong but it never works. Even when I use an example picture, it disconfigures it and takes the pieces apart and moves them all around in a weird way every time I try it. Do you know of any way to fix this? Thanks in advance
Very insightful! Could you tell me the difference between increasing iw (1 to 2 to 5 to 10000) and increasing the number of reference links (I think they both make the results have more similar features of the original)
It is actually fascinating to me how the art community of our generation is basically "feeding" AI by collecting data + crypted informations. Somehow, we are currently developing the foundations that will allow ultra comprehensive conversations with "AI". How mesmerizing..
Mesmerizing but scary. A lot of artists lost years of hard work and jobs because of AI.
@@blahblahbear1429 Art became cheap and democratic. For better or worse. I have sufficient talent to be an artist myself but never pursued this venue cause there have been too many artists already. I enjoy AI generators immensely.
@@brexitgreens watch many artists in the future, including children, lose hope and passion to try art because “AI just does it better.” It’s really fucking sad.
@@blahblahbear1429 You're right. There are new AI programs able to interpret any style. The sad part is realizing devs are literally stealing data from notorious platforms such as artstation or pixiv.. Anyone can copy the style of notorious artists like Stanley Lau 'artgerm' or even artstyles fom the most famous renaissance painters. I'm geniunely ok with it because ten thousand years from now, none of us will be around anyway but AI will remain somehow so...probably for centuries and beyond, who knows
@@blahblahbear1429 No, it's not. It democratizes art. It opens up the range of possibilities.
Is it possible to put an URL without touching the source and add only backround?
Is it possible to give the shape of the ref image a low weight. But the color a high weight? Because the clouds you made also changed shape based on the ref.
can you update / explain the VERSIONS and how this would apply to these techniques? (V5 was released this week march 2023) thanks!
Midjourney now analyses the images. Uploading 3 times exactly the same image as reference is prohibited in the rules. It can get your access revoked.
They have to differ from eachother.
thanks for your work, great job really. I've just tried the bot withe the 3 same images sequence, but Midjoiurney bot return this : One or more of your image prompts are the same image. Circumventing this filter to violate our rules may result in your access being revoked. -- is not normal?
Hello,
My free trial has ended, and I'm unable to sign up for the paid plan since I don't have a credit card and Stripe isn't currently accessible in Kenya. I only have PayPal, Please advise on how to resolve this.
Try using the one called Stable diffusion, it's free and works very similar :)
Hi!
I upload the images but when clicked, they do not open.
Nice tutorial. After you drag and drop, how do you send the image? I happens so quickly in the video. Thanks.
Thanks for watching! You drag and drop your image, once the image is loaded and you see it in your text box at the bottom just press Enter!
@@cyankatana Thanks. For anyone listening in, it seems the problem was the Safari browser. It works fine in Google Chrome.
@@gregfuller999 thanks for sharing this. Really good to know!
thank u
I tried this but mj removes the --iw parameter each time I click enter. Why?
Thank you for sharing, does anyone know, how to get the link on an iPad?
What's the benefit/reason for doing multiple prompts at once per Enter key, vs. seperating them out into different queries with multiple Enter keys?
Hi Kaizen! Not sure I understand your question. But if you are referring to the fact that I have many prompts in the text box - that was purely for demonstration on the screen, I ran each individually.
so i'm getting errors with --iw is there a specific version of MidJourney that you are using for this video... I am trying to use --v 4 with the --iw prompt
does not work with v4
@@rm-gi2mj Seems to work with version 3 ... ???
Great tutorial. I am watching it 4 months later and there are some changes, or maybe I am doing something wrong. When I try to prompt 3 of the same picture, I get a message that filters prevent that prompt for multiple same pictures.
Apparently you cannot use the same photo multiple times any more I got an error, "Duplicate images detected...
Could you give multiple image prompts for the same image different weights? 🤔
Great question. Yes you can do that by separating the images with :: and following that up with a number. Similar to text weights (have a tutorial on that too)
I tried this with 3 of the same image and I got this message using. Each was a unique link. Not sure what I didn wrong.
'Duplicate images detected
One or more of your image prompts are the same image. Circumventing this filter to violate our rules may result in your access being revoked.
Our image filter is currently a work in progress, and may have false positives.'
question: Can this work the same way for a full body picture of a character?
Example: a full body reference pic of Green Lantern Kilowog from DC comics
It could, although it will be a lot more difficult - the more details you want Midjourney to get right the harder it will be. My example was a very basic composition, thus the results were very similar. Give it a try!
lit!
🔥
Now I need a tutorial on how to add an image to discord. Yup, you guessed it, I'm "older". 🤨
You can just drag and drop the image if you are on your desktop. If on mobile, open the channel you want to send it to and then click "+" sign next to your text box at the bottom.
@@cyankatana Thanks.🙂
I tried to use multiple images and it flagged that i have used multiple duplicate images and that if I try to circumvent this I may be banned access from Midjourney. Where did I go wrong? All the images had different links. Also before that it said i cant give any weight in v4. not sure what that means
If you don't specify any words does it behave like -iw 9999 ?
Good question. I'm not really sure. I did try without any words and it was just a random mess of an image haha
I love you
"One or more of your image prompts are the same image. Circumventing this filter to violate our rules may result in your access being revoked. "
I got this warning when provided the same images with different links.
and also the --iw is not available in V4.
I am unable to open the image
Duplicate images are not allowed (anymore) it says..
damn learned so much, buy i used up all midjourney free credits and sad they will not give any more for free
Yeah, unfortunately you do need to Subscribe if you'll be using it seriously!
how can I use MidJourney in my channel?
This video explains it all - ua-cam.com/video/3Ok1K912A8Y/v-deo.html
@@cyankatana thank you
I lost you on send your image to discord... ?? Why do instructors say "simply" and "it's so easy"
I have no commands like /imagine . Just on direct message with midjourney bot
Interesting. I play around with pics of me for future album covers. And so far (3 weeks in) latest from today, following all suggestions, is, that I (a blond man middle long hair) created the black and white portrait of a dark haired woman. ::iw 5. That is... strange. You blond woman result is 100 x better.
Join the Discord in the description and I can try to help you out with that! No guarantees of success thought haha
Just tried the 3 images from of me from the same photo session, very similar angled shots, plus ::iw 1. The grid produces 1 Landscape and 3 abstract pictures. Humph. 🙂
Helpful video. Just what i was looking for to increase likeness of end result. But just received this message from MIDJOURNEY when i try to upload multiple same image like you suggested
👇👇
Duplicate images detected
One or more of your image prompts are the same image. Circumventing this filter to violate our rules may result in your access being revoked.
_______________
Suggestions? Thanks.
Weird. Why would they even care?
It's Jessica
Using the same image multiple times as a reference doesn't seem allowed anymore. It says "Duplicate images detected
One or more of your image prompts are the same image. Circumventing this filter to violate our rules may result in your access being revoked.
Our image filter is currently a work in progress, and may have false positives."
One or more of your image prompts of same duplicate image violates their rules may result in access being revoked.
Midjourney should begin making plans to raise their 25 free limit right away. As an alternative, AI technologies like Blue Willow will become more and more popular over time. Blue Willow, in my opinion, will be the one to take down MidJourney
you can just pay 10 a month its worth it
You didn't Include a seed number, so it's hard to tell how much was affected by the reference vs just a random seed difference
i think iw commands is not working anymore
Even with 5.1, it doesn't work very well with animals.
8:32 "platinum blonde" would work better
It would have been wonderful if you'd included how to upload an image to your Discord server. Great tutorial otherwise!
Thanks for your comment Katie! It is included, I simply dragged and dropped the images from my Downloads folder to Discord and hit Enter (that was my downloads folder you saw on the top right)
@@cyankatana I see that now. It was just me being befuddled as usual. :-) And thanks again for the reply and help. I got it to work! :-D
Can you add me to your server?
Not sure if anyone else is getting this result now:
Duplicate images detected
One or more of your image prompts are the same image. Circumventing this filter to violate our rules may result in your access being revoked.
Our image filter is currently a work in progress, and may have false positives.
yep, getting this
switch over to v3
@@sputtahsputtah6137 how?
Lol you made a NYYXXII.
Had to Google 🤷
stable diffusion is so much better at doing this.
Right click and nothing happens
Apparently using multiple images as a prompt is against their rules now
I've spent hours and hours, dozens of photos, trying to get art based on it, and it simply won't work. It works just fine with photos found online though. Videos like this are misleading, and are the cause of me signing up and finding out it was absolutely pointless.
Those nostrils lmao.
you are not an artist
Ure not either
Feel better now? Ah, you gotta love social __cking media.
@@flickwtchr much better, especially when i have geeks who think theyre hot shit like you because they can write combination of words into a text bar LOL
Yep AI isn’t art at all. It’s just images.
Like it or not this is the future of art no doubt AI is absolutely amazing.