I teach Midjourney full time. I have had almost a hundred one on one students. I cannot tell you how many people come to me with unrealistic expectations fed by misinformation on TH-cam. I really appreciate the accuracy of your videos so thank you for the hard work you have done to get this right.
95% of MJ photos look the same. the wes anderson style photos all have the same lifeless colors and composition, architecture all resemble each other despite the different characteristics and moving outside the bounds of these two things leads back to them one way or another.
Can I ask how you got into this man? (The teaching part that is!) I've been considering it but i have kept my work to myself till now but was intrigued to hear what you said as I used to teach at a high level on subjects for quite a few different industries - but didn't / haven't looked into the MJ one! If you reply - much appreciated! If not, that's ok too! :D
@@jzwadlo I started on gig sites it took time and it’s a tough place to make a go of it though. It took me a few months to get noticed enough to get work but I’m doing well at the moment. I also built my own website and made sure to be active on all social media. Twitter, TikTok, Instagram, Facebook, LinkedIn. I am working on a lot of projects. I work about 60hours a week including Saturday and Sunday. You know, pretty much building an empire 😂
People tend to focus on what can be done with a single prompt. But a lot of the more artistic and complex images are made using chains of prompts. The process goes like this: generate an image, then use the generated image as an image prompt, and so on and so on. This immediately pushes you out beyond the "stock" look of 99.9% of MJ images. It’s a workaround for the limitations of what a single prompt can do. It gives way more control over composition and the ability to insert things into the image that MJ sometimes struggles to incorporate. I've heard people are even using simple line sketches to set up the composition early in the process. I'd like to see more tutorials on this technique.
I've come to accept that, no matter how good your prompt is, MJ just isn't yet capable of delivering a very precise, detailed image that includes more than a few specific aspects in a single image. However, mixing images can deliver amazing results if you stage your prompts e.g prompt an amazing character in a single image, then prompt an amazing background/environment in a separate image, then combine them. Out of thousands of attempts trying everything to achieve a specific scene in the one prompt, I now save a lot of fast hours by combining images. And these are invariably some of my best images.
Still kind of new to this but this really intrigues me. Suppose you have a background image that you're pretty happy with him don't really want to change you just bring it in as part of the prompt but set it to zero weight?
@@FluxNomad678 You can certainly control the weight of individual images but, as Nolan will tell you, MJ is a lottery. If you have something really specific you want to achieve, MJ will make you work for it. And often in vain. There is a beauty in riding the wave and being open to being surprised by the unexpected. Just like Life. 🧶
HUGE thank you, I actually knew a lot of the tips but I completely forgot about the image weight, I will try it out right nowwwww thankssss ;) great video man: to the point and interesting too.
please try action sceane prompts for example: A giant car chases a biker through an apocalyptic city on fire for detailed image add cinematic backlight studiolight realastic verydetailed etc
meh. It's interesting for sure. But these systems are so deep and complex, I still have a lot to cover on Midjourney lol I'll get to other ai's eventually
Doesn't work for me when I paste in the image address. I get this error message: "Invalid parameter Cannot use --version 5.1 with only a single image prompt. Please add another image prompt, or a text prompt."
Yeah I think that means you need to include some words if you're trying to just use a picture, which is why I suggest using the /describe feature first to let Midjourney 'write' some prompts for you
This was extremely helpful! I was so hopeful this would give me the answers I need as I am trying to recreate a photograph as an illustration but MJ just seems to be so confused with the old man holding slices of pizza on his fishing rod LOL! I can't seem to get the prompt right, sigh
I just checked and I still see the Upload file option ... sometimes the 'use apps' shows up and sometimes 'create thread' is there too, but I think that has to do with the channel you're trying to create on ... anyways, I haven't found a place that didn't include 'upload file' so I'm not sure what's wrong on your end. Sorry I'm no help
It's a cool feature but not very accurate yet. I fed it a photo it had itself made and it came up with artist suggestions that make stuff that looks nothing like the style of the original picture.
It definitely seems to offer at least 1 weird suggestion every time I run it ... maybe that's by design? Like a long-shot chance that uses different words but still might work? I'm really not sure, just thinking out loud
@@FutureTechPilot Repeat the individual words that are being missed for example, A songbird sits on the shoulder shoulder shoulder shoulder shoulder shoulder of a shy shy shy shy shy shy shy blonde woman dressed in white
@@FutureTechPilot Thank you for keeping your Instructions simple and to the point. but I have a question how come the more you zoom out the darker the scene gets . its like its trying to fade to black ?
deceiving title. this video doesn't provide any accurate way to write midjourney prompts. it just a covers couple of examples, that's all. it doesn't provide any method to write accurate prompts.
What are you talking about? This video is about using /describe to show you which words help create a specific picture, and how you can use it to accurately write a prompt
@@marwich I used a picture I made myself. And by using the /describe feature, you're able to see how to accurately recreate it. Hope you have a good day
I teach Midjourney full time. I have had almost a hundred one on one students. I cannot tell you how many people come to me with unrealistic expectations fed by misinformation on TH-cam. I really appreciate the accuracy of your videos so thank you for the hard work you have done to get this right.
haha and I appreciate you taking the time to comment. Good luck on your teachings!
95% of MJ photos look the same. the wes anderson style photos all have the same lifeless colors and composition, architecture all resemble each other despite the different characteristics and moving outside the bounds of these two things leads back to them one way or another.
Can I ask how you got into this man? (The teaching part that is!) I've been considering it but i have kept my work to myself till now but was intrigued to hear what you said as I used to teach at a high level on subjects for quite a few different industries - but didn't / haven't looked into the MJ one! If you reply - much appreciated! If not, that's ok too! :D
@@jzwadlo I started on gig sites it took time and it’s a tough place to make a go of it though. It took me a few months to get noticed enough to get work but I’m doing well at the moment. I also built my own website and made sure to be active on all social media. Twitter, TikTok, Instagram, Facebook, LinkedIn. I am working on a lot of projects. I work about 60hours a week including Saturday and Sunday. You know, pretty much building an empire 😂
People tend to focus on what can be done with a single prompt. But a lot of the more artistic and complex images are made using chains of prompts. The process goes like this: generate an image, then use the generated image as an image prompt, and so on and so on. This immediately pushes you out beyond the "stock" look of 99.9% of MJ images. It’s a workaround for the limitations of what a single prompt can do. It gives way more control over composition and the ability to insert things into the image that MJ sometimes struggles to incorporate. I've heard people are even using simple line sketches to set up the composition early in the process. I'd like to see more tutorials on this technique.
Great suggestion! Thanks for pointing that out, I can start working on a video like that now
Thanks for the tip
I've come to accept that, no matter how good your prompt is, MJ just isn't yet capable of delivering a very precise, detailed image that includes more than a few specific aspects in a single image. However, mixing images can deliver amazing results if you stage your prompts e.g prompt an amazing character in a single image, then prompt an amazing background/environment in a separate image, then combine them. Out of thousands of attempts trying everything to achieve a specific scene in the one prompt, I now save a lot of fast hours by combining images. And these are invariably some of my best images.
Yeah I know exactly what you mean! I kinda touch on that in this video - th-cam.com/video/ICh9ose7tGs/w-d-xo.html
Still kind of new to this but this really intrigues me. Suppose you have a background image that you're pretty happy with him don't really want to change you just bring it in as part of the prompt but set it to zero weight?
@@FluxNomad678 You can certainly control the weight of individual images but, as Nolan will tell you, MJ is a lottery. If you have something really specific you want to achieve, MJ will make you work for it. And often in vain. There is a beauty in riding the wave and being open to being surprised by the unexpected. Just like Life. 🧶
HUGE thank you, I actually knew a lot of the tips but I completely forgot about the image weight, I will try it out right nowwwww thankssss ;) great video man: to the point and interesting too.
Thanks a lot for the feedback! I appreciate it and I'm glad you enjoy my type of videos. Cheers buddy
This was great! Gives me a better idea of how I can successfully create better prompts, thank you!
Happy to help!
Good one. Tried it and it worked very well.
Happy to hear that!
Thanks for the vid. I knew some of this already but the --iw was new to me! Much appreciated
haha happy to help, thanks for the comment!
Wow! Super awesome Tips! Thank you very much for sharing this! 👍
Cheers!
Nice! Great deep dive into /describe!
It's a little niche, so thanks for the support!
I love to create art with Midjourney AI. thanks for sharing informative tutorial.
And thank you for the comment!
Thanks!
Woah. Thanks!
Perfect 😊 thank's
Cheers!
please try action sceane prompts for example:
A giant car chases a biker through an apocalyptic city on fire
for detailed image add cinematic backlight studiolight realastic verydetailed etc
Sure! Thanks for the suggestion!
Is there a way to use /descrive on the site?
kind of. Upload an image. Scroll over it and you'll see a little 'i' button. Click that and you'll something similar to describe. Hope that helps!
Anything on leonardo?
meh. It's interesting for sure. But these systems are so deep and complex, I still have a lot to cover on Midjourney lol I'll get to other ai's eventually
Great - thanks
🤝
😻👍
🤝
Describe sucks. Generated prompts are not indicative of how a prompt should be written. Use it for brainstorming and nothing else.
haha yeah it's not perfect, but definitely helps you learn about different key words
👋
🤝
Doesn't work for me when I paste in the image address. I get this error message: "Invalid parameter
Cannot use --version 5.1 with only a single image prompt.
Please add another image prompt, or a text prompt."
Yeah I think that means you need to include some words if you're trying to just use a picture, which is why I suggest using the /describe feature first to let Midjourney 'write' some prompts for you
This was extremely helpful! I was so hopeful this would give me the answers I need as I am trying to recreate a photograph as an illustration but MJ just seems to be so confused with the old man holding slices of pizza on his fishing rod LOL! I can't seem to get the prompt right, sigh
haha yeah it's not as straight forward as we would like
Did they remove the Upload file option? I don't have it when I click +, only 'Create thread' and 'Use apps'
I just checked and I still see the Upload file option ... sometimes the 'use apps' shows up and sometimes 'create thread' is there too, but I think that has to do with the channel you're trying to create on ... anyways, I haven't found a place that didn't include 'upload file' so I'm not sure what's wrong on your end. Sorry I'm no help
@@FutureTechPilot I'm in a newcomer room. Maybe I need to dig around (and ask around) more. Thanks for responding!
thank you sir
🤝
It's a cool feature but not very accurate yet. I fed it a photo it had itself made and it came up with artist suggestions that make stuff that looks nothing like the style of the original picture.
It definitely seems to offer at least 1 weird suggestion every time I run it ... maybe that's by design? Like a long-shot chance that uses different words but still might work? I'm really not sure, just thinking out loud
Essentially: without the picture as a reference, you will not get something similar enough
Yeah, it does seem impossible to recreate complex images just through words alone
@@FutureTechPilot but same image with another object?How?
@@paulmoldovan9518 I think your best bet is to use something like Photoshop
@@FutureTechPilot just tried with a cat. Great result. Same prompt with link but this time a dog. Wow futuristic dog :)) totaly failed
That was very helpful thank u❤
Glad to hear!
I love these deep dives you do into the nuts and bolts of MJ. Would love to see you do one on sliders and cowbells.
haha I'm glad other people want to learn as well... and I use sliders for sure, but what do cowbells refer to?
@@FutureTechPilot Repeat the individual words that are being missed for example,
A songbird sits on the shoulder shoulder shoulder shoulder shoulder shoulder of a shy shy shy shy shy shy shy blonde woman dressed in white
Excellent MJ tutorial. Thanks. Liked and Subscribed.
Thanks a lot for the support!
Thanks for the plethora of info. It's already change my perspective on things!
haha well I'm glad to help, thanks for taking the time to comment!
I Just tried this on make Marvels Captain Carter in her superhero fightsuit and it worked like a dream thanks
I'm glad to hear that!
@@FutureTechPilot Thank you for keeping your Instructions simple and to the point. but I have a question how come the more you zoom out the darker the scene gets . its like its trying to fade to black ?
Awesome tips and highlights brah! 🎉
Cheers buddy!
great video ! peace
Cheers buddy!
Fantastic tips! 🎉👏🏻💯
I'm glad to hear that!
deceiving title. this video doesn't provide any accurate way to write midjourney prompts. it just a covers couple of examples, that's all. it doesn't provide any method to write accurate prompts.
What are you talking about? This video is about using /describe to show you which words help create a specific picture, and how you can use it to accurately write a prompt
@@FutureTechPilot that's not true. you just picked a prompt from someone else and tried it yourself. that's not an "how to" video
@@marwich I used a picture I made myself. And by using the /describe feature, you're able to see how to accurately recreate it. Hope you have a good day