Chiming in and echoing everyone’s comments - thank you so much Christian for eloquently articulating your process and for pushing things forward. I think I know things but then you come in with a new perspective. Really appreciate the work you put into educating and wanted to give you the encouragement to continue to do so. You are incredible.
Outstanding - you're a talented and empathetic presenter. The attention you pay to simplicity, brevity and details (such as how you show the prompts) is soooo worth it 🏆✔
Man, I love your channel so so much. It's helping me out a lot! Thank you so much for your time and effort to explain everything. Your explanations are so well structured it's so easy to reproduce everything in Midjourney.
Blend is pretty amazing, the fact that you have no control and the result is a surprise is just part of the fun. My favourite images ever are a result of blending.
Some great tips there. I tried Blend when it first was introduced. Wasn’t overly thrilled with it, but now will have to have a try using these methods.
Hey Christian, I just discovered your channel recently and it's my new favourite Midjourney resource! Great in-depth info, and really well explained.Thanks for everything you do.
Just found your channel. Been using MJ for about two months and learned a lot just in this video. Your teaching style works for me. Just subscribed to your channel, email list, and course waiting list. I'll snag that as soon as you release it.
Christian!! I really apreciate the effort you put in all this videos!! thank you very much!! i havent found a video teaching how to correct images (like misplaced hands in a dragon or three legged horse) do you have any video related to that!? thank you again for everything!!!!
Hey I just wanted to say I can see the effort you've put into your set and your look. All freshly shaven etc. It's noticeable and I've noticed the effort you've put in. Well done. It's looking great and more professional. 🎉 Keep up the great work!
Ironically, I've not really done anything in particular regarding the "set". I'm just in a different location right now and need to work with what I have 😅 But if it looks good, then that's great. Thanks 😁
This is great in not only showing the strengths but the weaknesses of ai art. In the loft architecture case I notice how frustrating it was for you to actually get what you wanted and how you ultimately had to give up and compromise. This could be for a number of reasons but is most likely because the two architecture styles are so contrasting; diffusion is bringing things into focus from a blur and so contrast gets lost.
I really like the last confirmation of the methods otherwise I was going to take the video to the beginning then watch again which is most of the time making me feel that I have a long learning curve 😢 and feel stressed...But you are really good at teaching and showing the new ways...This detail is my reason to subscribe and thumbs up to you 🎉 thanks
And people still say that it's "oh so easy to use Ai, and it's going to replace us". Look at how many steps this guy just had to do to get to the result he wanted. This is the "human" side, mixing and matching a bunch of techniques and methods to one or more AI tools. Really great video with lots of tricks and tips to achieve incredibly precise results. Thank you.
thx, really good tutorials! I would like to add some tipps from my experience: for the describe feature I can recommend trying all four prompts and check the results instead of just picking what you expect to fit best. Often its not the prompt that I think matches well that in the end gives me the closest results. And in your second method for the white staircase you could try to mention it twice in the prompt with the higher weight. Sometimes it helps to get closer to the result you had in mind, but not always ✌
Great content as always, Christian!! Another interesting thing is that we can also use the link of some of the images you got the description for and feed it into the merged prompts! There are really infinite ways to keep on producing images until we are satisfied with the results!
Great vid, thanks you, will try these techniques. The /describe is a new one for me may prove to be very game changing. I have a question about workflow if anyone would like to weigh in on it. I create event artwork for an underground house music event and have used Midjourney to create an recognisable brand from the art generated, however I have been using the blend command excessively with varied results. I was wondering as a workflow possibility how would creating a composition using various collaged elements created in photoshop work as a starting point for creating a more polished and improved version of that comp within Midjourney?
Very interesting. I think we're going to converge on a set of verbs (or operators) that you can then use to predictably get the results you want to get. Kind of like having an algebra for images.
Well done, Christian. There is, I believe, a typo in your subscriber count. Clearly, you should be at 4.36 million! Thanks for the excellent tutorials.
Ohhhh myyyy! Now I’m at work screaming and cussing at my phone in excitement and I look insane and I don’t care! 😂 this video helped me sooooo much! Thank you!!!
Excellent content. Is there any way to control the layering when combining images, e.g. image (1) is the window of a house and image (2) is the silhouette of a person. How to control the combination in a multiprompt that the person comes out behind the window instead of in front of the window? - Any advice appreciated. I tried experimenting with the sequence and weight of the multi-prompts, but the results are still random and seldom satisfactorily putting the person behind the window.
@@davidmichaelcomfort first picture shows my scene with random people. followed by the picture of my two consistent characters and the last picture shows the results
@@davidmichaelcomfort it was a very long journey to get these pictures. and this video gave me the key for these results. 1. describe each picture with midjourney 2. and the weighted :: gives the ability to add (in principle) >2 characters to the pictures.
@@christians.8963 Thanks! To clarify, you (1) generated first image with two random people, (2) two images with one consistent character each, (3) use describe to get text prompt for each of the three images, (4) combine all three prompts into a multi-prompt with different weighting. Is this correct? One tip I can share to get consistent characters is to use an actor/model/public figure’s name to get consistent character across images. E.g. “[model name] as a medieval prince”. This works much better than using image prompts or seeds or other techniques. I can get a consistent character in pretty much every image I create. It is two consistent characters in one image which is really challenging. Thanks again!
Blend is terrific however I was was getting a little disheartened with results ranging from "wow' to is 'that all you got?'. These are fantastic and a brilliant example of lateral thinking. Thank you.
I always like the way you explain it. Starting from pouring the idea into a prompt, then executing it into a picture, then how do you solve the problem if the resulting image is not what is expected, which part of the prompt must be adjusted so as to produce a final image that is satisfactory, close to what is expected, or even better. I learn a lot from your videos. Thanks!
Very clever, elegant, and useful. I am compelled to personally express my gratitude for this video among the diluge of AI content that I have devoured this year. Maraming salamat.
This is awesome. I have a super hero that I have been trying to make. He is a human with bird wings connected to his arms and body (like a bat but with feathers) that I have not been able to generate. I am going to try this tonight.
Idk. You could still use blend to get something similar to what you want and then have midjourney describe that, instead of having midjourney describe two different images and then combining them. I think the main distinction here is that blend is still extremely useful for combining two art styles and or just using it as a research tool to generate different ideas or to explore different concepts.
Funny thing is: That first alternative you just described is exactly what I show at the beginning of the video. Did you watch the whole thing? Also, don't take the headline too seriously. Of course you can still use /blend. I honestly have just found it to be rather "meh" recently.
I am just curious, what could be the relationship between MJ and architecture/interior design? I will appreciate an answer from you, a real master of AI MJ.
Ok. You can use MJ to create art..art is subjective, with no limits...architecture starts with a sketch, and finishes up with dimensions...so everything is very exact, based on numbers. I do not see MJ or AI, at this moment, being able to deal with all these numbers. Watching your video with the two interiors, how you transform the NY apartment using the med style, is still a sketch, an idea... you can have as many variants as you want; but is not a final product, as is a painting, a logo, or a business card, etc... So, AI is not yet ready to be applied in architecture. Am I right? Oh, Yes, if I understand well....is only '/imagine'! "Currently, we are witnessing designers and creative professionals use AI for concept generation purposes, but with time AI will become an important part of the overall design process."
I can not seem to prompt mid journey to create a flat ceiling. i do mid century modern houses with flat, or no slope, or level ceilings. whatever I add in, it does not do it. As soon as I tell it about a beam, it turns the beam into a cathedral ceiling or mono pitch.
Tolles Video. Aber es demonstriert wieder einmal schön, dass die KI zwar weiss, wie etwas AUSSIEHT, aber nicht, wie es FUNKTIONIERT. In diesem Sinne kann man die meisten gezeigten Wendeltreppen nicht verwenden, weil die eine Hälfte in die eine Richtung dreht und die andere Hälfte in die andere Richtung. Vielen Dank fürs Teilen. Viele Grüsse aus der Schweiz 🇨🇭
Jupp. Allerdings vergessen auch viele, dass die Technologie noch in den Kinderschuhen steckt. Gleichzeitig wird auch unterschätzt, wie schnell sich das alles verbessern wird.
Hey man, your content is truly amazing! I have a suggestion and request. For all the 3D artists out there, it would be interesting to see if Midjourney could create beautiful renders of interiors with an orthographic view. This way, it can be easily modeled based on the render. It would be great to have a bird's-eye view or any other suitable orthographic view of the environment as well. Thanks again!
You mean taking an existing image and asking Midjourney to create and orthographic view of the exact same image/space? You can't really do that because Midjourney isn't even aware of 3-dimensional space, let alone the ability to view the same space from various angles. It will always create a new image with each command. But maybe I'm misunderstanding?
@@TokenizedAI I understand, this is what I have been failing to do but what if there was a way to also generate an orthographic view of something mid journey generates? like the interior you show in your video
@@TokenizedAI Here are a few prompts that seem to be getting close but not quite: multiple views orthographic and prespective of interior decorating 3d room rendering, in the style of light white and dark brown, hard edge painter, dada-inspired constructions, subtle, earthy tones, light yellow and white, matte photo, simple designs --ar 253:75 --------- interior design with 2d virtual reality, in the style of dark white and dark beige, martiros saryan, dark brown and dark beige, loose paint application, kentaro miura, simple designs, octane render --ar 253:75 --------------------------- 4 schematic drawings of the military helicopter, in the style of hervé guibert, kodak vision3 250d 5207, clean and streamlined, dau al set, made of all of the above, govaert flinck, jaume plensa --ar 51:62
@@salemwillson9059 But that's my whole point. Midjourney isn't creating a "space" or "interior" and then deciding to display it from a particular perspective. From what you are saying sounds like you're assuming that Midjourney is arranging a particular room that you can then view from various angles. But that's not what it does. It's simply creating an image and best fits the description of the prompt. It doesn't actually understand each specific object, let alone the fact that it could a 3-dimensional space. It's doesn't even understand what a 3-dimensional space is. I could imagine that you might be able to create something that resembles an orthographic view. But it won't really understand what it is. At it won't be able to get both the orthographic and 3-dimensional view of the very same space. That's simply not how diffusion models work. At best, they'd have to create their own technology that is capable of infering a 3D object from a 2D image. That sort of stuff exists but it's far from perfect. I think the fact that MJ can produce such realistic-look images is misleading people to believe that it's somehow far more capable and intelligent than it actually is.
@@TokenizedAI I see what you are saying, but I have on multiple occasions managed to get multiple views or looks of the same room within the same midjourney generation. If I say !make me 4 images of the same room from different angles with different lighting conditions" its literally the exact same room from multiple angles. Please play with the prompts I gave and you will see what I mean. I am by no means a master, but If it can do multiple shots of the same room then there is no reason for it not to be able to create specific views?
can you please do a video of how can i use midjourney with my product that i want to put it in différents scene that looks like it been photographed prefectionly
Chiming in and echoing everyone’s comments - thank you so much Christian for eloquently articulating your process and for pushing things forward. I think I know things but then you come in with a new perspective. Really appreciate the work you put into educating and wanted to give you the encouragement to continue to do so. You are incredible.
Thank you so much for this comment 🙏🏻🤗
Great video, Christian, keep up the excellent work
You're a prompt scientist! The tests, experimentations and care in your videos is absolutely insane, thank yo so much for sharing your knowledge
Life is all about experiments 🙂
Outstanding - you're a talented and empathetic presenter. The attention you pay to simplicity, brevity and details (such as how you show the prompts) is soooo worth it 🏆✔
I try to present how I'd enjoy learning 🙂🙏🏻
The greatest video about Midjourney I've seen for a while. Finally, someone started to make content with real tips and tricks !
Thank you so much 🙏🏻
This tutorial is exactly what I needed! I was stuck using /blend and couldn’t move forward. Thank you!
Glad I could help! 🙌🏻
The best midjourney video I’ve seen, by far.
🙏🏻🤗
Man, I love your channel so so much. It's helping me out a lot! Thank you so much for your time and effort to explain everything. Your explanations are so well structured it's so easy to reproduce everything in Midjourney.
Thank you for confirming that my approach is appealing to some people 🙂
This was the best tutorial in mid journey that I have seen. I had no idea that there was a describe prompt total game changer. Thanks.
When you know, you know 😉
Blend is pretty amazing, the fact that you have no control and the result is a surprise is just part of the fun. My favourite images ever are a result of blending.
Yes, I partially agree. But it doesn't help when the image contains obvious flaws that aren't helpful for what I'm trying to create.
I use it with abstract portraits that I get from other platforms, and it is amazing.
You have become my go to person for all things Midjourney. Never would have thought of this method of blending on my own.
Thank you. I appreciate comments like these so much 🙏🏻🙂
You're channel, this video, just took my MJ fun and skills to a whole new level!
Awesome! 🙂
such a great video to understand how prompts work in midjourney! thank you for detailed and step by step explanation.
My stuff just increased in quality, thank you
Great to hear! 😀
Yoor content never dissapoints. Thank you!
Thank you 🤗
Some great tips there. I tried Blend when it first was introduced. Wasn’t overly thrilled with it, but now will have to have a try using these methods.
Hope you enjoy it!
Best MJ trick I've learned in a long time. Thank you, you're a genius 😊
Pleasure
Very Dope Alternative to just Blend. Your strategy is cool while allowing you to think outside of the box. Rhanks!
Glad you like it!
these were great methods, that I started playing around with! thank you!
Great stuff Christian. This kind of brainstorming and working through examples really helps me increase my Midjourney skills
Brainstorming is key to progress 🙂
Hey Christian, I just discovered your channel recently and it's my new favourite Midjourney resource! Great in-depth info, and really well explained.Thanks for everything you do.
Thanks so much 🙏🏻
Just found your channel. Been using MJ for about two months and learned a lot just in this video. Your teaching style works for me. Just subscribed to your channel, email list, and course waiting list. I'll snag that as soon as you release it.
Awesome and welcome! 🤗
Great video! Appreciate the process driven design.
Great tutorial!! This opens up literally endless possibilities! Thanks!
one of the best videos I've seen about mid journey, prompt engineering and everything in between
you teach to think, not in ctrl c ctrl v
Thinking is such an underrated skill nowadays. ngl! 🤣
Here is the prompt knowledge I need to manipulate Midjourney to produce what I imagine. Finally!
Christian!! I really apreciate the effort you put in all this videos!! thank you very much!! i havent found a video teaching how to correct images (like misplaced hands in a dragon or three legged horse) do you have any video related to that!? thank you again for everything!!!!
Wait for Inpainting
The rabbit hole gets deeper and deeper. 🙂 Taking the art of synthography to higher levels. Great stuff 👍
Rabbit holes are fun and dangerous
Great work. Real cutting edge application combined with creative problem solving. Impressive. TY.
Thank you very much! 🙂
Hey I just wanted to say I can see the effort you've put into your set and your look. All freshly shaven etc. It's noticeable and I've noticed the effort you've put in. Well done. It's looking great and more professional. 🎉
Keep up the great work!
Ironically, I've not really done anything in particular regarding the "set". I'm just in a different location right now and need to work with what I have 😅 But if it looks good, then that's great. Thanks 😁
Thank you! professionalism and perfection!
Thank you 🙏🏻
One of the best Midjourney tutorials I've watched hands down......I feel like Picasso now. Lol
Hahaha, let's make some art!
i adore your way of explaining things
Thank you 🤗
Ottimo, è una modalità che stavo già sperimentando anch'io appena MJ ha proposto /describe, ciao e grazie
This was packed with useful information. Thanks!
Glad it was helpful!
This is great in not only showing the strengths but the weaknesses of ai art. In the loft architecture case I notice how frustrating it was for you to actually get what you wanted and how you ultimately had to give up and compromise. This could be for a number of reasons but is most likely because the two architecture styles are so contrasting; diffusion is bringing things into focus from a blur and so contrast gets lost.
Yeah, indeed. My frustrations aren't even adequately conveyed through the video 🤣
I really like the last confirmation of the methods otherwise I was going to take the video to the beginning then watch again which is most of the time making me feel that I have a long learning curve 😢 and feel stressed...But you are really good at teaching and showing the new ways...This detail is my reason to subscribe and thumbs up to you 🎉 thanks
Thank you so much for your feedback. I just try to explain it the way I would want to learn it 🙂
And people still say that it's "oh so easy to use Ai, and it's going to replace us". Look at how many steps this guy just had to do to get to the result he wanted. This is the "human" side, mixing and matching a bunch of techniques and methods to one or more AI tools. Really great video with lots of tricks and tips to achieve incredibly precise results. Thank you.
Yeah, creating art with AI is super easy. Just like you can get a randomly awesome photo by accident. But doing it with intent is a different story.
Excellent video. Lost of info without fluff. Saved for later. Subscribed and liked. Cheers
Lots of ***
Thank you 🙏🏻
Yo that Mr. Clean reference was everything! 😂your videos are 🔥🔥
Haha, glad someone noticed 😂
As always BRILLIANT! So helpful. Thanks so much for taking the time to create this, and then sharing it :)
Glad it was helpful!
Hey, I use /blend to mix animals together to new hybrids and get great results. I guess it depends on what you're blending
The claim was intentionally hyperbolic. Not be taken too literally.
Of course, it hybrid creatures are what you want, then go for it 🙂
thx, really good tutorials!
I would like to add some tipps from my experience: for the describe feature I can recommend trying all four prompts and check the results instead of just picking what you expect to fit best. Often its not the prompt that I think matches well that in the end gives me the closest results.
And in your second method for the white staircase you could try to mention it twice in the prompt with the higher weight. Sometimes it helps to get closer to the result you had in mind, but not always ✌
Yes, I've done this too. I actually mention this. But it takes too much time for this type of video format.
Great content as always, Christian!! Another interesting thing is that we can also use the link of some of the images you got the description for and feed it into the merged prompts! There are really infinite ways to keep on producing images until we are satisfied with the results!
Yup, I agree 🙂
Soon this will all be considered part of a possibility space you can scroll through with sliders in real time.
Wow, you rock man!! This is very useful to me as a web designer, thanks for sharing!! You deserve five stars ⭐️
Glad it was helpful! And thank you for the tip. Gonna try this on web designs too 🙂
Amazing yet again Christian, thanks again for sharing your understanding of Midjourney...
My pleasure!
Thank you for all this information. Definitely going to add these to my workflow in MidJourney. 🤘🏼
Go for it!
Amazing stuff! Thank you for sharing and teaching all these creative workflows. Kudos and keep the videos coming!
Thanks, will do!
learned so much and took notes, thank you
Glad you learned something 🙂
Thank you so much! It's mind expanding.
WHOA! I have like 3 more tools in my toolbelt! Thank you!
You're welcome! 🙂
Great vid, thanks you, will try these techniques. The /describe is a new one for me may prove to be very game changing. I have a question about workflow if anyone would like to weigh in on it. I create event artwork for an underground house music event and have used Midjourney to create an recognisable brand from the art generated, however I have been using the blend command excessively with varied results. I was wondering as a workflow possibility how would creating a composition using various collaged elements created in photoshop work as a starting point for creating a more polished and improved version of that comp within Midjourney?
Thanks Christian ! Great Video ! Even Carla Caruso would be proud of it ! 😀
Haha 😅
This was really helpful!! This is going to help make some interesting art in midjourney! Thanks so much!!
Glad it was helpful!
Midjourney turns us all into frustrated magazine feature editors.
lol, why?
Thanks for this insight on how to use MJ to get better than average results.
You bet!
This is a rather interesting approach that you have come up with love the permutations that you were doing to create tiny refinements
Thank you! I thought so too, which is why I shared it 😁
Very interesting. I think we're going to converge on a set of verbs (or operators) that you can then use to predictably get the results you want to get. Kind of like having an algebra for images.
Yes, potentially
Well done, Christian. There is, I believe, a typo in your subscriber count. Clearly, you should be at 4.36 million! Thanks for the excellent tutorials.
Haha, I wish 😅
Thank you for helping us to become better creators! Your amazing! Bravo!😊
You are so welcome!
Very nice pro tips! 🎉
Glad you think so!
Another great and inspiring video. Thank you so much! ❤
Very interesting! Thank you, Christian!
Great idea. Thanks for the insights in this video
My pleasure!
Ohhhh myyyy! Now I’m at work screaming and cussing at my phone in excitement and I look insane and I don’t care! 😂 this video helped me sooooo much! Thank you!!!
Hahahahaha, made my day 🤣
This is absolute Gold.
Thanks
Super creative and way to think outside the box!
Thank you 🤗
this is an amazing tutorial. thank you!
You're welcome!
Great content, Chris! Thanx so much
Pleasure 🙏🏻
Excellent content. Is there any way to control the layering when combining images, e.g. image (1) is the window of a house and image (2) is the silhouette of a person. How to control the combination in a multiprompt that the person comes out behind the window instead of in front of the window? - Any advice appreciated. I tried experimenting with the sequence and weight of the multi-prompts, but the results are still random and seldom satisfactorily putting the person behind the window.
You are brilliant, thanks to your video I am now able to create two different consistent characters in one image with a consistent background!
I’m curious how you got two different consistent characters in one image. MJ always seems to blend the characters together. Thanks I’m advance
Fantastic!
@@davidmichaelcomfort first picture shows my scene with random people. followed by the picture of my two consistent characters and the last picture shows the results
@@davidmichaelcomfort it was a very long journey to get these pictures. and this video gave me the key for these results.
1. describe each picture with midjourney
2. and the weighted :: gives the ability to add (in principle) >2 characters to the pictures.
@@christians.8963 Thanks! To clarify, you (1) generated first image with two random people, (2) two images with one consistent character each, (3) use describe to get text prompt for each of the three images, (4) combine all three prompts into a multi-prompt with different weighting. Is this correct?
One tip I can share to get consistent characters is to use an actor/model/public figure’s name to get consistent character across images. E.g. “[model name] as a medieval prince”. This works much better than using image prompts or seeds or other techniques. I can get a consistent character in pretty much every image I create. It is two consistent characters in one image which is really challenging.
Thanks again!
Thanks Christian - excellent content! 😀👍🏽
Thanks 🙂
Blend is terrific however I was was getting a little disheartened with results ranging from "wow' to is 'that all you got?'.
These are fantastic and a brilliant example of lateral thinking. Thank you.
Yeah, just blending images is a bit limited, unfortunately. Hence why this approach provides an alternative pathway.
It was great! thanks for the video!
Glad you enjoyed it!
Awesome video! Thanks for the tips.👏
How to blend too images with consistent styling, it should not modify the final image, it should just perfom color match.
Thank you 🙏 Such good tips and pedagogically put forward too. You're a 🌟
I appreciate that!
This was really cool. Took a break from the consistent character series to watch this one! I will have to try that photobashing technique sometime!
I haven't even thought about how this might impact character creation 🙂
I always like the way you explain it. Starting from pouring the idea into a prompt, then executing it into a picture, then how do you solve the problem if the resulting image is not what is expected, which part of the prompt must be adjusted so as to produce a final image that is satisfactory, close to what is expected, or even better. I learn a lot from your videos. Thanks!
Life is all about problem-solving. Unfortunately, most people think there's a silver bullet for everything.
Really excellent! Thanks for the brilliantly delivered, useful info!
Glad you enjoyed it!
Thanks, great vid. Very clear explainations and great examples.
As always, an excellent video. Of course, you're getting the hang of it! . Thanks Christian.
🤗
Very clever, elegant, and useful.
I am compelled to personally express my gratitude for this video among the diluge of AI content that I have devoured this year. Maraming salamat.
My pleasure 🙂
You are the best of the best! Thanks a lot for sharing
Wonderful sharing, thank you!
This is awesome. I have a super hero that I have been trying to make. He is a human with bird wings connected to his arms and body (like a bat but with feathers) that I have not been able to generate. I am going to try this tonight.
Good luck! 🙂
Amazing stuff! Thanks for sharing your knowledge.
My pleasure!
Idk. You could still use blend to get something similar to what you want and then have midjourney describe that, instead of having midjourney describe two different images and then combining them. I think the main distinction here is that blend is still extremely useful for combining two art styles and or just using it as a research tool to generate different ideas or to explore different concepts.
Funny thing is: That first alternative you just described is exactly what I show at the beginning of the video. Did you watch the whole thing?
Also, don't take the headline too seriously. Of course you can still use /blend. I honestly have just found it to be rather "meh" recently.
interesting video with quite some functions that slipped my research but 10:18 you call this "not malformed"?
Yes, still malformed. But not immediately obvious as with the others. MJ still struggles with stairs.
I am just curious, what could be the relationship between MJ and architecture/interior design? I will appreciate an answer from you, a real master of AI MJ.
Not sure I understand the question 🤔
Ok. You can use MJ to create art..art is subjective, with no limits...architecture starts with a sketch, and finishes up with dimensions...so everything is very exact, based on numbers. I do not see MJ or AI, at this moment, being able to deal with all these numbers. Watching your video with the two interiors, how you transform the NY apartment using the med style, is still a sketch, an idea... you can have as many variants as you want; but is not a final product, as is a painting, a logo, or a business card, etc... So, AI is not yet ready to be applied in architecture. Am I right? Oh, Yes, if I understand well....is only '/imagine'! "Currently, we are witnessing designers and creative professionals use AI for concept generation purposes, but with time AI will become an important part of the overall design process."
eres el mejor,, graciasssss por tu generosidad al enseñarnos
I can not seem to prompt mid journey to create a flat ceiling. i do mid century modern houses with flat, or no slope, or level ceilings. whatever I add in, it does not do it. As soon as I tell it about a beam, it turns the beam into a cathedral ceiling or mono pitch.
Is there a very specific technical term for that sort of ceiling or architecture? You might be using ambiguous words.
Tolles Video. Aber es demonstriert wieder einmal schön, dass die KI zwar weiss, wie etwas AUSSIEHT, aber nicht, wie es FUNKTIONIERT. In diesem Sinne kann man die meisten gezeigten Wendeltreppen nicht verwenden, weil die eine Hälfte in die eine Richtung dreht und die andere Hälfte in die andere Richtung.
Vielen Dank fürs Teilen.
Viele Grüsse aus der Schweiz 🇨🇭
Jupp. Allerdings vergessen auch viele, dass die Technologie noch in den Kinderschuhen steckt. Gleichzeitig wird auch unterschätzt, wie schnell sich das alles verbessern wird.
You are a hero::2 Thank you::1
Thanks 😁
Thanks for the great tips! Love your work.
Thanks 🙂
I'm very grateful to you for sharing your fantastic, easy to follow videos and thank you also for the great resources you've provided
Super glad to hear that 😁
Incredible, thank you for the ideas
great insights, thank you 🙏
Hey man, your content is truly amazing! I have a suggestion and request. For all the 3D artists out there, it would be interesting to see if Midjourney could create beautiful renders of interiors with an orthographic view. This way, it can be easily modeled based on the render. It would be great to have a bird's-eye view or any other suitable orthographic view of the environment as well. Thanks again!
You mean taking an existing image and asking Midjourney to create and orthographic view of the exact same image/space?
You can't really do that because Midjourney isn't even aware of 3-dimensional space, let alone the ability to view the same space from various angles. It will always create a new image with each command.
But maybe I'm misunderstanding?
@@TokenizedAI I understand, this is what I have been failing to do but what if there was a way to also generate an orthographic view of something mid journey generates? like the interior you show in your video
@@TokenizedAI Here are a few prompts that seem to be getting close but not quite: multiple views orthographic and prespective of interior decorating 3d room rendering, in the style of light white and dark brown, hard edge painter, dada-inspired constructions, subtle, earthy tones, light yellow and white, matte photo, simple designs --ar 253:75 --------- interior design with 2d virtual reality, in the style of dark white and dark beige, martiros saryan, dark brown and dark beige, loose paint application, kentaro miura, simple designs, octane render --ar 253:75 --------------------------- 4 schematic drawings of the military helicopter, in the style of hervé guibert, kodak vision3 250d 5207, clean and streamlined, dau al set, made of all of the above, govaert flinck, jaume plensa --ar 51:62
@@salemwillson9059 But that's my whole point. Midjourney isn't creating a "space" or "interior" and then deciding to display it from a particular perspective. From what you are saying sounds like you're assuming that Midjourney is arranging a particular room that you can then view from various angles. But that's not what it does.
It's simply creating an image and best fits the description of the prompt. It doesn't actually understand each specific object, let alone the fact that it could a 3-dimensional space. It's doesn't even understand what a 3-dimensional space is.
I could imagine that you might be able to create something that resembles an orthographic view. But it won't really understand what it is. At it won't be able to get both the orthographic and 3-dimensional view of the very same space. That's simply not how diffusion models work.
At best, they'd have to create their own technology that is capable of infering a 3D object from a 2D image. That sort of stuff exists but it's far from perfect.
I think the fact that MJ can produce such realistic-look images is misleading people to believe that it's somehow far more capable and intelligent than it actually is.
@@TokenizedAI I see what you are saying, but I have on multiple occasions managed to get multiple views or looks of the same room within the same midjourney generation. If I say !make me 4 images of the same room from different angles with different lighting conditions" its literally the exact same room from multiple angles. Please play with the prompts I gave and you will see what I mean. I am by no means a master, but If it can do multiple shots of the same room then there is no reason for it not to be able to create specific views?
can you please do a video of how can i use midjourney with my product that i want to put it in différents scene that looks like it been photographed prefectionly
You'll need to use Photoshop for that. Watch my video on how I created a Gin brand.