Hi, thanks for tips. I have one for you! You do not need to use so many words in one Prompt. That's not how LLMs works. What happens is the image gets "muddied." You are having MJ looks at TRILLIONS and TRILLIONS and TRILLIONS of permutations. You don't want to do that. My images are 100% photo-realistic, using 1/10 of the number of prompts you use. Just a tip. The more prompts, the WORSE the outcome. You can leave off many. It's how LLM work. Try the experiment, think you may be blown away with the results. :-)
In fact i was wandering why in the video he uses promps that contain parts like: "Use the multi-prompt "car::photorealistic::rust" with a prompt weighting of "rust" to emphasize the aged textures and worn out look of the car" Aren't these guidelines for the user to apply to the previous part of the prompt!? LOL
You are correct. MJ parses the text to the AI using 72 tokens. The more words that are used in the description the less emphasis each word has because it's effectively trying to allocate the 72 tokens across a greater number of words. That's why shorter commands often produce better results.
Wow! That's an amazing tip resource! Thanks for putting that together. I've only recently found MJ and am dumbfounded by how powerful a tool it is. I'm a writer and poet and (poor) artist who has had to outsource art in the past, but this can enable me to create my own books, my own vision, without having to rely on an intermediary.
Dude nice! I got some really nice images for my website using some of these. I love the emoji idea. So often we want to convey emotions or concepts and an emoji seems like a really powerful way to concentrate that.
@@aiadvantage Awww thank you! You're so engaging...I'm like too excited to sleep now hahaha...currently working on my projects as we speak & feeling so inspired! :D
I think this year we will see lots of amazing AI tools that will make our lives easier. Those who can use these AI tools smartly will be the gainers. Recently I created amazing images with Blue Willow
ok but... why in some prompts we are telling midjourney things like: "Use the multi-prompt "car::photorealistic::rust" with a prompt weighting of "rust" to emphasize the aged textures and worn out look of the car" Aren't these guidelines for the user? Why wasn't the weighting directly applied in the previous part of the prompt!? I'm confused...
Hi ! Thank you! Cool! I want to get such a picture, what does a person who falls from the roof of a skyscraper see, or a person who fell from the window of a tall building, what would he see at that moment (first-person view)?- do you know how to get such pictures ?
I do agree. Some bad ducking pulled the voice down for no obvious technical reasons [maybe the background music triggered it]. Valuabel content but irritating sound quality
What UA-camrs Don't Tell You About AI Art: 1.Approval required: To sell on most major Print on Demand (POD) platforms, you need to be accepted as a seller or artist beforehand, which can take a significant amount of time depending on demand. 2.High competition: There is an immense amount of competition, and AI-generated art is not as profitable as it may seem, since the market is saturated and professional designers are earning their living in the same space. 3.Clickbait videos: UA-camrs who create videos with thumbnails displaying specific numbers like '$7748.39' are selling a dream and making money off your clicks. 4.Passive income potential: be real here, the market is saturated hope of selling anything, or else your pieces will be lost among the countless others. 5. Competition: and lack of promotion on side of platforms, i have uploaded 1000 full fledges designs with perfectly optimized SEO. Guess what i have zero sales. why, because anyone can do it nowadays. 6. Research: Plus, they skip the research part which is the very first step and the foundation of all your actions.
I just want to know how to even start to use Midjourney lol no matter what I do it never makes an image for me and always says there is to many users using it try again next time, i been trying for weeks now and cant do anything ever or am i missing something?
Bro why didn't you give the download link. I have been trying to get into it for an hour but could. I wanted to remove unwanted cells to make it personalized, any help?
hmm you can train your own version of stable diffusion for sure but that's no easy task. if you want a free alternative look into "stable difussion XL" or "Leonardo AI"
I have a question. How can I select and if it is possible to do so at all to choose part of an image and combine it with a part of another image? Say I like the human version in 1 version and the animal version in another image. Right now I am thinking photoshop lol
@@aiadvantage thank you. I don't understand how to send the bot images as a message part. Would you mind explaining in more detail please? My understanding is to load the images somewhere and put in the prompt. Also, how to specify which part of each image to use in order to combine?
The compressor settings on the vocal for this video are so bad it's distracting. I think the release time on the compressor is way too long and it's causing a pumping effect which noticeably modulates perceived volume.
you are right. funny thing is i actually had it off but there was a second compressor in the background of my soundboard that still did this. recording completely clean now so not an issue anymore
sorry mate ! just a dumm as question. we are trying to add a baseball cap to our character and we get just garbage from midjourney. do you know how to add a baseball cap?
why is it called AI-"Art", its generated pictures, where is there an art in it ? or is it just to make people think they are an artist by writing some words and wait an hour to select a picture which doesn't look too much crappy ?
Hey thanks heaps for your efforts! ❤ Can anyone offer some advice on why I would not have got any images at all, even waiting hours? I was so pumped to sign up to discord, got in the newbies section, entered some pretty straight forward prompts just to get my feet wet, but got no images back at all. 😢 i tried several times, changed the prompts, even changed the chat room, even tried sending a message if anyone could read my prompts to see if I was doing something wrong (disappointingly got no reply in 3 requests- so I didn’t feel the love of a community I was expecting) So I was pretty disappointed. Lost my mojo… Any ideas why?? 🤔
or, for example, such pictures where they are shown in the first person: what would a child see at the moment when he is abducted?- please help me how to get such pictures
It's all too clean, no grit, no reality. Noone should compare real human art to ai art. There is no comparision..ai art is stereotypically "good" to people who know nothing about art and have bad taste as well. It can barely be classified as stock illustration.
I son’t know man if I look at the character designs, patterns or fantasy its pretty damn excellent. I see your point with many other categories but remember this is day 1.
Hi, thanks for tips. I have one for you! You do not need to use so many words in one Prompt. That's not how LLMs works. What happens is the image gets "muddied." You are having MJ looks at TRILLIONS and TRILLIONS and TRILLIONS of permutations. You don't want to do that. My images are 100% photo-realistic, using 1/10 of the number of prompts you use. Just a tip. The more prompts, the WORSE the outcome. You can leave off many. It's how LLM work. Try the experiment, think you may be blown away with the results. :-)
You got a guide to follow?
In fact i was wandering why in the video he uses promps that contain parts like:
"Use the multi-prompt "car::photorealistic::rust" with a prompt weighting of "rust" to emphasize the aged textures and worn out look of the car"
Aren't these guidelines for the user to apply to the previous part of the prompt!? LOL
Agreed! Long worded prompts = beginner level prompting
You are correct. MJ parses the text to the AI using 72 tokens. The more words that are used in the description the less emphasis each word has because it's effectively trying to allocate the 72 tokens across a greater number of words. That's why shorter commands often produce better results.
Wow! That's an amazing tip resource! Thanks for putting that together. I've only recently found MJ and am dumbfounded by how powerful a tool it is. I'm a writer and poet and (poor) artist who has had to outsource art in the past, but this can enable me to create my own books, my own vision, without having to rely on an intermediary.
What an incredibly amazing video and prompts document you have put together. This is so generous and a game changer...thank you so much!
Mind-blowing 🤯 Thanks you so much for sharing this, Igor 🙏
Usuallz the youtuber is the one using that word. Good to see it within a happy comment :)
@@aiadvantagesorry 🤭
Man this is legit gold. Thank you for doing what you are doing. Subscribed today and can't seeing that changing anytime soon lol
Dude nice! I got some really nice images for my website using some of these. I love the emoji idea. So often we want to convey emotions or concepts and an emoji seems like a really powerful way to concentrate that.
I love it! I didn't know at least 4 or five of them, super informative, thank you.
Gratulation zu mittlerweile über 100.000 Abonnenten. Tolle Arbeit, top Erfolg. Weiter so!
LOVE this tutorial video-- so informative and beautifully done! Had to subscribe. Thank you SO much! :))
Love to hear it! Welcome on board
@@aiadvantage Awww thank you! You're so engaging...I'm like too excited to sleep now hahaha...currently working on my projects as we speak & feeling so inspired! :D
I think this year we will see lots of amazing AI tools that will make our lives easier. Those who can use these AI tools smartly will be the gainers. Recently I created amazing images with Blue Willow
Excellent! Thank you so much!
Best Original AI Content out there.
Can't wait to see the exactly this PDF on Copy pasted by 20 other channels.
Exactly what I was needing! Thanks man
Over the top delivery young man. Highly recommend you blast out an agency.
You are the Prompt King.
ok but... why in some prompts we are telling midjourney things like:
"Use the multi-prompt "car::photorealistic::rust" with a prompt weighting of "rust" to emphasize the aged textures and worn out look of the car"
Aren't these guidelines for the user? Why wasn't the weighting directly applied in the previous part of the prompt!? I'm confused...
very educational and helpful. Keep going and don't stop uploading. you have my support
great video nd the advice is amazing! Can't wait to implement these into my new designs
Ayyy, superb video!!👏🏿👏🏿🙌🏿 I’m subscribed, time to⚙️!!🙏🏿
I feel MJ 4 is more for artistic results while MJ5 is giving great photo and cinematic and rendered results
Very inspirational! Thanks a lot!
HI Igor - how do we load an image to Midjourney so it can use it as a basis for an image? Thanks!! Cary
You are doing everything so right, well done man!
Thank you so much, this is very helpful. Much appreciated!!!
This is super helpful, thank you so so much!!!
Thanks for the work on this...appreciated.
How did NOBODY notice the three legged ballerina yet? I gave it 24 hours but not a single comment 😂🤷♂️ #mjv4thinga
This was fantastic! Thank you!
thanks for the resource :)
Thank you, very useful info. !
These are great! Thank you!
Totally wrapped in this and your awesome generosity. Have joined your discord. Love your work
Thank you a lot for what you share for community
Simply awesome. You, Sir are a living legend!
You make it easier to understand some other people make it much more complicated
Awesome. You're saving me so much time.
Thanks!
Awesome video Thank you!
This is amazing, tysm! ♥
Thanks for sharing!
Thanks
Wow. Thank you!!
Many Thanks!
Thanks so much, Igor. Great help.
This is my go to sheet right now. Abuse it, enjoy it and thanks for the comment
Very cool, thank you!
Díky!
thank you for the spreadsheet. Appreciate your work . sub'd
I can import a portrait of someone but when I use mid journey the result looks nothing like them...How do I retain their face as it is...?
Thank you
Brilliant 👍😎👍
Hi ! Thank you! Cool!
I want to get such a picture, what does a person who falls from the roof of a skyscraper see, or a person who fell from the window of a tall building, what would he see at that moment (first-person view)?- do you know how to get such pictures ?
Pure gold! Thanks a lot!
Thanks so much for the info. One question: can the art you create be copyrighted?
Tough question as this still is a greyzone. Midjourney claims you get commercial usage rights when you purchase one of their plans that's all we know
Great video, thanks :) But the sound quality was some how bad.
I do agree. Some bad ducking pulled the voice down for no obvious technical reasons [maybe the background music triggered it]. Valuabel content but irritating sound quality
Thank you ❤️
I just saw this video I was looking for how to do newborns digit backdrops
So first thing now I need to buy this but how can I do this 😅
What UA-camrs Don't Tell You About AI Art:
1.Approval required: To sell on most major Print on Demand (POD) platforms, you need to be accepted as a seller or artist beforehand, which can take a significant amount of time depending on demand.
2.High competition: There is an immense amount of competition, and AI-generated art is not as profitable as it may seem, since the market is saturated and professional designers are earning their living in the same space.
3.Clickbait videos: UA-camrs who create videos with thumbnails displaying specific numbers like '$7748.39' are selling a dream and making money off your clicks.
4.Passive income potential: be real here, the market is saturated hope of selling anything, or else your pieces will be lost among the countless others.
5. Competition: and lack of promotion on side of platforms, i have uploaded 1000 full fledges designs with perfectly optimized SEO. Guess what i have zero sales. why, because anyone can do it nowadays.
6. Research: Plus, they skip the research part which is the very first step and the foundation of all your actions.
I just want to know how to even start to use Midjourney lol no matter what I do it never makes an image for me and always says there is to many users using it try again next time, i been trying for weeks now and cant do anything ever or am i missing something?
Lots of quality nuggets in there. Mahalo!
Thank u so much 😊
Did you say Midjourney is not Free anymore? I haven't signed up yet, but if I do, will I be required to purchase a plan to use it?
It's not free.
They stopped offering 25 images for free around 10 days ago
Cheapest plan is $10 mth
Bro why didn't you give the download link. I have been trying to get into it for an hour but could. I wanted to remove unwanted cells to make it personalized, any help?
Sorry for asking more, since midjourney not giving free trail... Id like to know if there is chance can we create our own midjourney?
hmm you can train your own version of stable diffusion for sure but that's no easy task. if you want a free alternative look into "stable difussion XL" or "Leonardo AI"
Chefs kiss and hats off to ya
How do you create an image greater than 1024x1024 72 DPI in MidJourney?
use an upscaling website like : upscale.media to double the resolution
I have a question. How can I select and if it is possible to do so at all to choose part of an image and combine it with a part of another image? Say I like the human version in 1 version and the animal version in another image. Right now I am thinking photoshop lol
you can send bot images as a message and then copy the 2 links and include them in your /imagine prompt!
@@aiadvantage thank you. I don't understand how to send the bot images as a message part. Would you mind explaining in more detail please? My understanding is to load the images somewhere and put in the prompt. Also, how to specify which part of each image to use in order to combine?
LOVE YOU CHANNEL
Amazing!!!
Great 👍 This is amazing
How do I create a Pixar-type cartoon with midjourney?
Great techniques!
Mega! Danke🎉
The compressor settings on the vocal for this video are so bad it's distracting. I think the release time on the compressor is way too long and it's causing a pumping effect which noticeably modulates perceived volume.
you are right. funny thing is i actually had it off but there was a second compressor in the background of my soundboard that still did this. recording completely clean now so not an issue anymore
Great video ! 👍👏
sorry mate !
just a dumm as question.
we are trying to add a baseball cap to our character and we get just garbage from midjourney.
do you know how to add a baseball cap?
why is it called AI-"Art", its generated pictures, where is there an art in it ?
or is it just to make people think they are an artist by writing some words and wait an hour to select a picture which doesn't look too much crappy ?
The second seems more true...99% is not art at all.
amazing thank you.
Good work man, +1 for you.
Hey thanks heaps for your efforts! ❤
Can anyone offer some advice on why I would not have got any images at all, even waiting hours?
I was so pumped to sign up to discord, got in the newbies section, entered some pretty straight forward prompts just to get my feet wet, but got no images back at all. 😢 i tried several times, changed the prompts, even changed the chat room, even tried sending a message if anyone could read my prompts to see if I was doing something wrong (disappointingly got no reply in 3 requests- so I didn’t feel the love of a community I was expecting)
So I was pretty disappointed. Lost my mojo…
Any ideas why?? 🤔
Are you typing in exactly this: /imagine then space then whatever you want to create
Hi, the link you provided isnot opening. :(
Amazing channel
Awesome
or, for example, such pictures where they are shown in the first person: what would a child see at the moment when he is abducted?- please help me how to get such pictures
I really need midjourney
You're awesome!
:) ♥️
Why is it not allowing me to save or download the Google Sheet prompts file?
Copy and paste the URL. 😉
Great - thanks a lot. PS - your mic sound is not that good.
great vid 🤠
wirestock doesnt accept ai images though? or am i wrong?
Nice man
373rd... Thanks... Great that I found your YT channel
your audio volume is fluctuating. are you a compressor?
good stuff... sub'd
Good stuff…
but what I know is you can't sell AI generated Art specially on websites like shutterstock!?
shitterstock you cant. the ones i mentioned you can
Best channel
Dude, take it easy on the audio compression.
🙏🏼
Well now they know :)
Where i can get free ai what makes nice pictures?
Try Nightcafe, stable diffusion or dalle-2
It's all too clean, no grit, no reality. Noone should compare real human art to ai art. There is no comparision..ai art is stereotypically "good" to people who know nothing about art and have bad taste as well. It can barely be classified as stock illustration.
I son’t know man if I look at the character designs, patterns or fantasy its pretty damn excellent. I see your point with many other categories but remember this is day 1.