serious question: how often do you run the commands in the example to get a result youre looking for? i mean... when i enter the exact same prompt like show in the video i do not get any results close to them. they look so different. is it only for showing how the image should look or should my results close to be your ones??
Well there are endless possibilities and a same word combination can result in so many things. My hope is that if I show you as many different subject matters, they will resemble at least something. And sometimes it also shows if something doesn't really work. I don't know if the seeds are currently still broken, but I have been able to replicate my prompts again when using the seed number 777. I have heard that it may sometimes happen that using the relaxed mode and fast mode (and also possibly turbo mode) can generate different images, because those generations are using different infrastructure in the background. The talking or moving images don't have seed numbers, because those were random. But could you specify what you are attempting to do, which prompt did you try to enter? I will try to see if I still get the same result or something different.
@@thaeyne Yeah, it is really hard to explain, because it is AI and not human. I mean, your images in the videos look always stunning, awesome, crystal clear and sharp and i would love to get descent results like shown. I wouldnt say my results are looking bad. As i said, it is so hard to explain, because using the macro command and getting an awesome looking closeup of a beautiful colored eye will not go in my head, because when i entered the prompt 1:1 i got a close up of a bee (not bad looking) but why a bee and not an eye, or a fruit, or a flower etc. It is so much trial and error, but how to get an image that you have in mind, which you can modify on your needs withouth getting completely different results?
@@Lomaxxx Well every character in your prompt actually matters. I even tested that depending on if you use --relax, --fast or --turbo, you will get slightly different results with the prompt "/imagine prompt: airbrush art Macro --seed 777 --v 5.2" I'm sorry I don't have any better answers, it is really hard and frustrating sometimes getting the exact results you want. It is a lot of trial and error. My guess about the different results in the macro are that all of them are kind of close-ups and the airbrush art adds some extra sharpness to it I think. So as Macro is not usually something totally specific, Midjourney kind of takes over and dreams up something random and not specific like bee or eye. Here are some /describe results of the image that I think you wanted to get (sorry if I got it wrong): 1⃣ a woman's eye that has some colorful paint and flowers on it, in the style of realistic hyper-detailed rendering, azure and amber, object portraiture specialist, dark orange and yellow, igor zenin, fish-eye lens, highly detailed figures 2⃣ an illustration of a woman's eye with a painted background, in the style of mixes realistic and fantastical elements, azure and amber, daz3d, vibrant murals, intricate illustrations, dark orange and yellow, highly detailed illustrations 3⃣ an artwork of a girl using flowers to decorate her eyes, in the style of photorealistic eye, sky-blue and amber, uhd image, highly detailed illustrations, stephen shortridge, colorful abstractions, colorful, eye-catching compositions 4⃣ a woman's eye with colourful flowers on it, in the style of realistic and hyper-detailed renderings, azure and amber, action painter, dark orange and yellow
You keep wondering what a zoom out would do, so why not do a video on that? Use a style for the first image, then add a word to the zoom, like "Fantasy-Inspirered art" and for zoom replace that with "Sci-fi" and see how Midjourney combines the two.
Amazing video, new suscriber. Started today with midjourney and maybe is stupid question but for example if I want design a landscape with this neo-traditionaljapanese stlye. How should I ask for it? /imagine prompt landscape, neo-traditionaljapanese ... and add elements i wish in it? Thanks a lot
Thank you, and yeah, that's basically it. My trick is to not overcomplicate things and just state what you want as shortly as possible and just be surprised at the amazing stuff that comes out.
The aspect ratios can depend upon how you’re looking at the image (e.g. from a phone, within discord, in a blog), where sometimes portrait works really well.
I’ve been trying to create an image of a tidal wave made of grass without luck. I’ve don’t around 50 tries using different wording, but have not been able to achieve it. Any ideas how I could create something like that? Love your vids. Thanks
Some of these images are quite striking I wonder what journey can do if giving the prompt triple I'm curious to see what images it would produce please do a video on this subject
Thank you! I used D-ID for the talking animations and the head movements and blinking, and Leiapix for the depth animations, the links are in the description field for all the tools I used.
@@thaeyne I am sorry i am hitting in this, but do you know how i can get animate an animal like this? D-ID only recognize human faces. That would be bonkers
@@Lomaxxx Yes it would be, but unfortunately it doesn't seem to recognize animal faces. I tried too with no luck. You could also try HeyGen, it seemed to recognize kind of a monster face I had. Didn't try with animals though.
![[UA] NAVI проти G2 Esports | Blast Premier Fall Final](http://i.ytimg.com/vi/QWlIm4FGESQ/mqdefault.jpg)
Thank you for uploading this video
serious question: how often do you run the commands in the example to get a result youre looking for? i mean... when i enter the exact same prompt like show in the video i do not get any results close to them. they look so different. is it only for showing how the image should look or should my results close to be your ones??
Well there are endless possibilities and a same word combination can result in so many things. My hope is that if I show you as many different subject matters, they will resemble at least something. And sometimes it also shows if something doesn't really work.
I don't know if the seeds are currently still broken, but I have been able to replicate my prompts again when using the seed number 777. I have heard that it may sometimes happen that using the relaxed mode and fast mode (and also possibly turbo mode) can generate different images, because those generations are using different infrastructure in the background. The talking or moving images don't have seed numbers, because those were random.
But could you specify what you are attempting to do, which prompt did you try to enter? I will try to see if I still get the same result or something different.
@@thaeyne Yeah, it is really hard to explain, because it is AI and not human. I mean, your images in the videos look always stunning, awesome, crystal clear and sharp and i would love to get descent results like shown. I wouldnt say my results are looking bad. As i said, it is so hard to explain, because using the macro command and getting an awesome looking closeup of a beautiful colored eye will not go in my head, because when i entered the prompt 1:1 i got a close up of a bee (not bad looking) but why a bee and not an eye, or a fruit, or a flower etc. It is so much trial and error, but how to get an image that you have in mind, which you can modify on your needs withouth getting completely different results?
@@Lomaxxx Well every character in your prompt actually matters. I even tested that depending on if you use --relax, --fast or --turbo, you will get slightly different results with the prompt "/imagine prompt: airbrush art Macro --seed 777 --v 5.2"
I'm sorry I don't have any better answers, it is really hard and frustrating sometimes getting the exact results you want. It is a lot of trial and error. My guess about the different results in the macro are that all of them are kind of close-ups and the airbrush art adds some extra sharpness to it I think. So as Macro is not usually something totally specific, Midjourney kind of takes over and dreams up something random and not specific like bee or eye.
Here are some /describe results of the image that I think you wanted to get (sorry if I got it wrong):
1⃣ a woman's eye that has some colorful paint and flowers on it, in the style of realistic hyper-detailed rendering, azure and amber, object portraiture specialist, dark orange and yellow, igor zenin, fish-eye lens, highly detailed figures
2⃣ an illustration of a woman's eye with a painted background, in the style of mixes realistic and fantastical elements, azure and amber, daz3d, vibrant murals, intricate illustrations, dark orange and yellow, highly detailed illustrations
3⃣ an artwork of a girl using flowers to decorate her eyes, in the style of photorealistic eye, sky-blue and amber, uhd image, highly detailed illustrations, stephen shortridge, colorful abstractions, colorful, eye-catching compositions
4⃣ a woman's eye with colourful flowers on it, in the style of realistic and hyper-detailed renderings, azure and amber, action painter, dark orange and yellow
Thank you for doing these kind of videos - they are quite informative and helpful, and I like them!!
You keep wondering what a zoom out would do, so why not do a video on that? Use a style for the first image, then add a word to the zoom, like "Fantasy-Inspirered art" and for zoom replace that with "Sci-fi" and see how Midjourney combines the two.
fantastic work, I would like to know if I can use the images you generate as references for my digital sculpture practice.
Thank you, and of course!
Great video as always! I really enjoyed it ❤
midjourney + did + leia
I'm really liking the talking animations. Good job!
Amazing video, new suscriber. Started today with midjourney and maybe is stupid question but for example if I want design a landscape with this neo-traditionaljapanese stlye. How should I ask for it?
/imagine prompt landscape, neo-traditionaljapanese ... and add elements i wish in it?
Thanks a lot
Thank you, and yeah, that's basically it. My trick is to not overcomplicate things and just state what you want as shortly as possible and just be surprised at the amazing stuff that comes out.
@@thaeyne thanks for the fast and clear answer. One question, I'm trying right now this. And looks our images are bit dark. How u balance this? Thanks
The aspect ratios can depend upon how you’re looking at the image (e.g. from a phone, within discord, in a blog), where sometimes portrait works really well.
I’ve been trying to create an image of a tidal wave made of grass without luck. I’ve don’t around 50 tries using different wording, but have not been able to achieve it. Any ideas how I could create something like that? Love your vids. Thanks
Some of these images are quite striking I wonder what journey can do if giving the prompt triple I'm curious to see what images it would produce please do a video on this subject
Really informative video! Thank you.
Yeah ! descriptions are back (me too). Kisses from France !
Appreciate your wonderful support to the community.
Thank you for this amazing collection of AI arts.
Love the animation, please tell us what tools you used. Great video and got excellenct inspiration
Thank you! I used D-ID for the talking animations and the head movements and blinking, and Leiapix for the depth animations, the links are in the description field for all the tools I used.
@@thaeyne I am sorry i am hitting in this, but do you know how i can get animate an animal like this? D-ID only recognize human faces. That would be bonkers
@@Lomaxxx Yes it would be, but unfortunately it doesn't seem to recognize animal faces. I tried too with no luck. You could also try HeyGen, it seemed to recognize kind of a monster face I had. Didn't try with animals though.
hello from Croatia and thanks!
Cool images 👍
Coulda been cut in half as it got boring after a while.
Machines are always overkill.