Midjourney - (the secret) to training Midjourney AI?

Is it possible to train Midjourney? so you can create the same character from different angles and perspectives. This is how to do it. The secret to training Midjourney AI, stable diffusion tutorial. And how to do character design with the help of Midjourney, DALL-E and Photoshop.
SUBSCRIBE
▶ handle: / @levendestreg
▶ You can subscribe to our channel here: kzbin.info...
▶ Read more: levendestreg.dk/en
▶ Download cheatsheet: levendestreg.dk/en/how-to-tra...
-------------------------------
Links
▶ Documentation: midjourney.gitbook.io/docs/us...
▶ Midourney's website: midjourney.com/home
▶ Once logged in click this link to see your images: www.midjourney.com/app/
▶ sketchfab.com/tags/head
▶ • DREAMBOOTH: Train MULT...
▶ monstermash.zone/
▶ huggingface.co/spaces/pharma/...
-------------------------------
This video is created on the Wacom Cintiq Pro 32 with Macbook Pro, Atem Mini Pro, Midjourney, DALL-E, Premiere Pro, After Effects and Photoshop.
00:00:00
Can you train Midjourney? I mean train it so that you can create the same character in different poses and from different angles? And if that’s really possible - why aren’t everybody else doing it?
00:00:19
And a lot of you wonderful people have asked about the possibility of somehow training Midjourney in what results you want. Or training it in different characters - or styles that you want to create.
00:00:56
Remember there are links in the description below among other things to a FREE CHEAT SHEET where I share some of my prompts with you.
00:01:25
What AI is the best for my needs? What AI can I get the best results with? And most importantly how can I train AI to create the character or style that I want to create?
00:02:09
But is that possible to achieve that with Midjourney? Well yes and no.
00:02:20
As the guidelines for Midjourney state - you cannot feed images directly into Midjourney like that due to concerns about community public content.
00:02:30
Instead Midjoueny lets you use images as inspiration - the img2img feature, usually along with text, to guide the generation of an image.
00:02:50
And as I explained in this video about character design
00:03:09
First let’s see if we can get Midjourney to give us the cropping and content we want
- the right part of the image, so to speak.
00:03:53
I’m going to give you seven different shots to choose from. There are a lot more. You just need to google. But for now, we’ll take a look at seven.
00:04:14
Now first we have the Full shot. Then the Medium full shot, Cowboy shot, Medium shot, Medium close-up, Closeup and Extreme close up. And then you have to choose your angle too.
00:04:54
So now you know the basics of what to ask Midjourney for. The trick here is to actually mention in what angle you see the face in - for instance. And I’ve had good results with mentioning details on the eyes too, so Midjourney doesn’t mess up the eyes.
00:06:05
To get the remix settings switched on you simply write /settings - and toggl on remix.
00:06:48
The term seed sets the seed for an image. So, when you use the term seed, it means that Midjourney will use the same noise or diffusion to create your image from.
00:07:04
And using seed can sometimes help keep things more steady and easier to replicate when trying to generate a similar prompt again or getting the same sort of image.
00:07:15
But - it can also be used for creating the same character in different poses, from different angles and perspectives.
00:07:53
Now how do you find the seed for the image you want then? Well, you add a reaction to the image you want to get the seed from. And you do this by right-clicking on the image.
Then you press “add reaction” and then “other reaction”. And here you just start to write env - and that’ll bring up the envelope.
00:08:14
Now Midjourney will send you a private message. And here you get the seed number.
00:08:47
And as you can tell - all of a sudden I have a character to work with. It’s magic.
00:08:55
So, what is same seed? You use the term -sameseed to affects all images of the resulting grid in the same way.
00:09:13
But if you use same seed - the four images will use the same slice. I can you use DALL-E’s paint out tool and Photoshop to do corrections.
00:09:29
If you want to learn more about DALL-E and the paint-out tool well, I’ve got an upcoming episode on that topic.
00:10:16
Because is there an easy way to turn your design into 3D models? The fast answer? Yes - and no! So, of course Adobe has some amazing tool to create 3D models.
00:10:31
And in general 3D has a steep learning curve. But of course there are always a shortcut or two that you can take.
00:10:41
So if you go to monstermash.zone/ and upload a 2D image - you can actually turn it into a 3D model - in a fairly simple way.

[@Neteroh]

I don't know how to start thanking you for this, me and a friend spent days without sleep trying to start a series and we didn't know how to keep our prompts consistent, we had a lot of beautiful characters and scenes that we didn't know how to redesign later or re-used and now with this trick we're done more sure of how to start the production, really thank you very much! It is a bit difficult to find good documentation of V4 and you have given it to us on a silver platter!! 🤩🤩😁😁

[@LevendeStreg]

Uhhh, I would love to see some of the stuff you create. Feel free to contact me on Discord. I’m @LevendeStreg#4449 - if you’re interested in sharing some thoughts! And thank you for the kind comment🙌

[@generalawareness101]

FRIEND, and I. Never ever put yourself first in an English sentence.

[@BrunoAraujo-po2lm]

I've noticed your channel went through a boom of growth when you started talking about Midjourney, I just want to make sure you know that your background as an illustrator and your enthusiasm with AI art generators are resulting on a HUMONGOUS help to all of us interested in AI art generators, PLEASE keep going because you are GOOD and bring such relevant points of view about the subject! And PLEASE give us more cheat cheets, it is very helpful. And of course THANK YOU SO MUCH!

[@LevendeStreg]

Wow @Bruno Araújo 🙏 thank you so much for your kind comment. It really means a lot to me, because I'm really putting lots of time and effort in every video. So, I'm glad you like the content and my input. So THANK YOU🙌

[@liquidmind]

@@LevendeStreg great job indeed

[@wisefoolman]

You are going to blow up with such high quality content like that. Developing AI art workflow is a major skillset to develop in so many industries. I'll be following closely what you put out and use it in my job.

[@LevendeStreg]

@@wisefoolman Thank you kindly - I really appreciate it. It takes a lot of time to keep up to date with the development right now. Everything is happening fast and progress is quick. It's really exciting to learn about AI right now! So doing my best☺

[@newfineart]

Just found your channel. Love the level of explaining you go to help us understand what is happening with the extra functionality within Midjourney.

[@LevendeStreg]

Thank you @XNY 1, I really appreciate it. And glad you like it🙌

[@IntiArtDesigns]

Hi, i just wanted to let you know i really like your videos. They are informative and easy to watch, and you bring a great energy to your presentation style which is concise and straight forward. The video and camera quality is good, the audio is crisp. Just well done all around. Keep it up and i think your subscribers will grow quickly. Thanks for the great content.

[@LevendeStreg]

Thank so much, @Paper2Gold for taking the time to comment🙏. I put in a huge amount of time and effort in the video, so good to know that it works and that people notice - and that you like the content 🙌

[@NoCodeFilmmaker]

Wow, great content and sorely under the radar. You should definitely be getting way more exposure. Great content and fantastic explanations. Keep up the awesome uploads 🙌💯🙏

[@LevendeStreg]

Wow thank you kindly, @NoCode Filmmaker 🙏 I really appreciate it! And I’m glad you like it!

[@lastmanstanding80]

Thanks for your tips and tutorials. Great channel, was surprised to see your sub count. You deserve more.

[@LevendeStreg]

Thank you kindly @Last Man Standing 🙏 I really appreciate that! And glad you enjoyed it.

[@Reinaldo0022]

Your tips are wonderful, thank you so much for sharing your content! I'm a 3d character artist and I started using midjourney to generate game character ideas recently, your tips are helping me a lot to learn about midjourney!

[@LevendeStreg]

Thank you @Reinaldo Di Pietro - that’s very kind of you 🙏 glad I could inspire you.

[@freestylelee216]

I was trying to finish the video before I commented but I couldn't help myself. I'm only like 5 minutes into the video and I already feel enlightened. Your enthusiasm, knowledge, and concise delivery are magnificent! Thanks so much.

[@LevendeStreg]

Thank you so much. I really appreciate it 🙌

[@woopee1677]

Thank you for your video! Really appreciate the work and effort you are putting into this.

[@LevendeStreg]

Oh thank you @Woo Pee, that really means a lot to me - thanks for returning to watch!🙏

[@minav.9323]

Congrats on your 1000k! ;) Thanks for the great video!

[@LevendeStreg]

Thank you so much @NinaV I really appreciate it🙌

[@loneranger4441]

Love the way you explain. Accidentally found your channel. Subscribed! 😊

[@LevendeStreg]

Thank you @Lone Ranger - I really appreciate it! 🙌

[@achliscantplay4202]

Levende, you are THE most helpful teacher on Midjourney that I have found. THANK YOU! Everything is compact, immediately applicable (!) and easy to remember. Even though it is not a professional requirement for me, and just a hobby, I was able to make so many cool images for my family, friends and students! (Including a super-weird request of "a fluffy cupcake in a magical forest").

[@LevendeStreg]

Thanks again @Achils Can't Play. I really appreciate you taking the time to comment and giving me the feedback. I put in a lot of time and effort in the videos. I love doing them. And it really helps for the motivation that people like you - like what I put out there. 🙏

[@soundalchemy2586]

Love your energy. Thank you for this great video.

[@LevendeStreg]

Thank you kindly @DavidRahming - I really appreciate it🙌

[@AlexandreRangel]

Great content, thank you Levende!

[@LevendeStreg]

Thank you kindly @AlexandreRangel. I really appreciate it🙌

[@ggelin2120]

I had to subscribe after watching this video. Thank you so much for this video and the cheatt sheet PDF!

[@LevendeStreg]

You very welcome, G Gein! I’m glad you liked it. And thank you for subscribing. I really appreciate it!

[@wabalabadubdub]

New subscriber here, You’re amazing! Thank you for the detailed information❤

[@LevendeStreg]

Thank you kindly @wabalabadubdub (and Love your hair). Thank you for taking time to comment🙏

[@troncalnegrete479]

Thank you very much you are very generous with your knowledge so far one the best tutorial about midjourney.

[@LevendeStreg]

Thank you for the kind comment, @Troncal Negrete. I really appreciate it🙌

[@DivDax]

Really like your videos! Thanks for sharing.

[@LevendeStreg]

Thank you so much. I really appreciate it! 🙏

[@hasigoki]

Damn. Love what you are doing. Sooooo helpful!!!

[@LevendeStreg]

I’m glad you think so @hasigoki2.072 - and glad I could help! 🙌

[@Aurapink]

Awesome! That was what exactly i needed ❤❤❤

[@LevendeStreg]

I’m glad you found it useful. Thank you so much, Aura. I really appreciate it 🙌

[@blurtam188]

Loving the content!!

[@LevendeStreg]

Thank you @Blur Tam. I appreciate it🙌

[@TheMiejoe777]

Thank you, this is super helpful!

[@LevendeStreg]

Glad you liked it @Miejoe, thank you for taking the time to comment🙌

[@TheMiejoe777]

No, thank *you* for making this video! :)

[@achliscantplay4202]

And a request for future tutorial - syntax, like what punctuation does, if anything, like: , + :: etc. I regularly see those in examples, but was unable to get a clear answer from people who use them yet, or to find clear info in official sources (maybe, I am not looking where I should be? Again, not an artist or programmer, so these could be obvious for anybody but me) 😅

[@LevendeStreg]

Got it @Achilis Can't Play. That's a good request. Makes a lot of sense. You can actually read about it, if you Google really hard on it - but I'll get that in a future episode. Thanks for your feedback again🙌

[@liquidmind]

you can read about that on MJ user manual, these works just like they are "," comma separator, + is not a know commando for me, but ppl use it to INCLUDE something, "::"" are weights. you give more emphasis to that word::3 for example.

[@LevendeStreg]

@@liquidmind Yup that is correct as @liquidmind says. Look for Advanced Text Weights for example on this page: midjourney.gitbook.io/docs/user-manual - I'm gonna do a summary on it in the upcoming video!

[@achliscantplay4202]

@@liquidmind Thanks mate, I appreciate that you are trying to help, but i have below median IQ and need to see actual examples and what they do to actually understand how to use things. People learn differently. I hope you trust that any sane person would actually try to do research and google things first, and only ask for help when they genuinely need it :) Levende's tutorials are what does it for me. Walls of text without example of usage and example of results cause me angry tears and throwing expensive tablets at dented walls :)

[@liquidmind]

@@achliscantplay4202 I have HaDD so I get you..... Wait until she makes one. I will help if she needs

[@stephangoirand3034]

Is it possible to train Midjourney with an object like a specific chair for example, so that it can recreate it correctly in any scenario?

[@LevendeStreg]

No. Not yet. But since it’s built on a stable diffusion engine - hopefully it will become possible in the future! 🙌

[@maiwyck]

Great and useful video! When you use “same seed” how do you format it with the prompt? I may have missed it but I didn’t see any visual example in the video of this.

[@LevendeStreg]

Thank you kindly @maiwyck🙌 I really appreciate it. What do you mean with format? "--sameseed" is the term - but you cannot use --sameseed with v4 of Midjourney yet, as far as I know. You can read the manual of the terms here: midjourney.gitbook.io/docs/user-manual - hope that answers your question. Otherwise let me know. And feel free to contact me on Discord - my username is @LevendeStreg#4449

[@liquidmind]

Just a note, stable diffusion you do not "feed it 10 images, and it makes art"... You actually create a MODEL with those images, a CHECKPOINT MODEL that THEN you use with MANY stable diffusion repos... Those models are trained on dream booth or textual inversion.... Unless you are referring to something else, or maybe you mean IMAGE TO IMAGE section in which you actually change images or make MORE of the same, without needing a custom model......

[@LevendeStreg]

Thank you @liquidmind for clarifying. I've seen some tutorials on it. I think we mean the same thing. But you are of course way more accurate in your description. 🙌

[@liquidmind]

@@LevendeStreg anyway, yes, people do not use --SEED that much, and its one of the best ways to keep coherence. So in a way of speaking as you said, MJ can be trained in a way, specially on v3 you get better results the more re-rolls you do, and now remixing with v4 makes it even more artistic. GREAT VIDEO indeed. I see many ppl looking for tutorials and I AM NOT good at that.

[@6ix1life]

Excellent video!! 👏👏

[@LevendeStreg]

Thank you kindly, so glad you liked it🙏

[@ozguraltay]

Hey Maria, thanks for the useful content. Camera shots and angles are very useful tips. Nonetheless, they don't give the exact positions of the character that you want to create. I've seen that some other AI artists solved this by using a very basic 3d mannequin. In that case, you should use img2img method. I'm using Stable Diffusion with AUTOMATIC1111's web UI, and I would suggest to check the new "depth 2 Mask" plugin for the img2img renderings. (It can be load from the extensions tab easily) Cheers.

[@LevendeStreg]

Thank you so much for this information @ozguraltay, that is so helpful. I will definitely look into that. Thank you for sharing you knowledge🙏

[@ozguraltay]

@@LevendeStreg you 're welcome.

[@igorvagun1984]

amazing information. thank you

[@LevendeStreg]

Thank you @Igor Vagun, glad you like it!🙏

[@Steger13]

Game changer.. We can make our own comics now😁 new sub.. I need to watch all your vdos.. Thanks

[@LevendeStreg]

Thank you so much @Steger - I really appreciate it! 🙌

[@GreyFoxRL]

Wow, this is great stuff! I am so glad I found you. :) This specific video tutorial is what I have been looking for. Thank you so much for doing this. I just have one short question. I may have missed something from the start, but do you mind telling me what did the image look like with the seed number 17648 that you used all through the video from the beginning? That would help us to understand how it all actually got started. Again, thank you for making your helpful videos. :)

[@LevendeStreg]

Hi there. Thank you so much. I really appreciate it- and thanks for watching. Found the image. But not sure how to share it with you. I can DM you on Discord?

[@samuelnaesen1176]

Haha I started doing that a few weeks ago. The angles I mean 😄 interesting. Brilliant tutorial btw

[@LevendeStreg]

Great, thanks for your comment, @samuelnaesen1176. Then you figured it out the same time as me I guess. It came up because I was doing a description for one of our illustrators, and then I tried adding it in a prompt.

[@harrygoldhagen2732]

Thanks for an excellent, enthusiastic tutorial! Unfortunately, I got a message from Midjourney that version 4 is not compatible with sameseed. Do you have a workaround?

[@LevendeStreg]

Thank you so much! I really appreciate it. And yes, for sameseed you need to use v3. 🙌

[@ze333]

Thanks for your video, it is really revealing. I still think that puppeting or keep the same character doing different animations, is very hard, specially when you have 2 characters and want them to interact with eachother...

[@LevendeStreg]

Of course you are totally right. Maybe check out stable diffusion. I’ve done a lot of videos on that. Thanks for watching🙌

[@IUSTITA]

You are a GEM!!!!

[@LevendeStreg]

Thank you kindly🙌

[@tripios]

hello thanks for the tips i have a question.. when im finding the seed what prompt im using after?? like if i have / imagine ""a man sitting on a table and eating"" what the prompt must be after?? / imagine ""a man raise up from a table - - seed 5456"" ?? or / imagine ""a man sitting on a table and eating, the man raise up from the table- - seed 5456?? ""

[@LevendeStreg]

That is actually a really great question, @tripios - thank you for asking it 🙌. I should probably explain this in an upcoming episode. So an example of a prompt could look like this: /imagine prompt: comic book style, style by xxx, a man with long dark hair sitting at a table looking at the camera while eating, detailed eyes, wide-shot, full body, --seed 5456 --ar 3:2 --test

[@adielynch7234]

can you teach Midjourney what something is with other photos if it does not appear to know? For example, Im trying to create some Pitch and putt golf holes. This is a type of golf that is played mainly in Ireland and Uk and is different than the standard golf hole in that each of the 18 golf holes on the course are very short and between 30 to 90 meters long. Midjourney does not appear to know this style of the game. Is it possible to teach it the term "pitch and putt" with references by uploading photos of different pitch and putt courses? what commands or prompts should be used to teach it what something is and teach it the parameters of the subject? tks

[@LevendeStreg]

No sorry. As it is right now, you cannot train or input learning like that into Midjourney as a user. You need Stable Diffusion for that!

[@LevendeStreg]

By the way, there might be a hack to "teaching" Midjourney something. You cannot teach it. But you can tweak your outcome. The new /describe prompt might be worth to look at for you. This way you can input images of pitch and putt golf holes that look correctly. And then you can see, what its' description of that is. Then you can tweak your prompt. I'll try to look into it and see, if I can make a video on it.

[@yeastydynasty]

Hi, I didn't see any other videos on your page covering this but do you have a workflow for putting two midjourney comic book characters together in one image?

[@LevendeStreg]

Oh that's a great request. No I haven't got a video on that. But I'll see what I can do. Normally I would try Stable Diffusion for that, though.

[@yeastydynasty]

@@LevendeStreg Ah okay. I used photoshop to comp the two characters together into a scene but the lighting for the characters didn't match the lighting in the background.

[@LevendeStreg]

@@yeastydynasty I would look at 3D and set up your figures there. But maybe check out Stable Diffusion. There are some crazy possibilities - like you ask for. See this video for example: kzbin.info/www/bejne/eZqqkKlsZsh_rqs

[@TheAbulletAway]

Great video and thank you. Was curious about seeds, are they global? Meaning, if I get a seed from one an image I create, can I send that seed info to someone else and they can use it?

[@LevendeStreg]

Yes, that is correct!

[@thaikytran3221]

Thank you very much for your clear instruction. Can you edit your link of cheat sheet, I can't download from your link above. Thanks once again and hope to recieve more your share.

[@LevendeStreg]

THank you kindly for watching and commenting. Here is the new link: levendestreg.dk/en/how-to-train-midjourney-ai/🙌

[@OrchestralMusicMidJourneyArt]

Good tips to help the newbies out there.

[@LevendeStreg]

Thank you kindly, @OrchestralMusicMidJourneyArt - I really appreciate it! Glad you approve! 🙌

[@TheCephalus]

Great vid Lavende. I just want to fact check you a little. Yes, AI can read your mind. the next thing in the diffusion development is fMRI to image diffusion. There is already a working prototype and could possibly get integrated into the next generation of augmented virtual reality glasses.

[@LevendeStreg]

Of course you’re right @TheCephalus - but in Midjourney’s version right now you still have to prompt correctly.😊

[@charlesneely]

My friend I have been able to recreate very painful childhood trauma in my past using AI you know even if I'm following trying to describe certain things once I hit quit generate I'm looking at s*** like I had time travel you know I recreated picture me as a little child in Mississippi and there I am wearing shorts and barefooted I never said anything about it and yes I did like to wear shorts a lot I swear man I could not tell it from a real photograph you know that he put everything in Black and White because that's what I remember things as back then I wouldn't see a color TV until 1972 when I got sent to the penitentiary so I think it reads the impulses in your eyes from your webcam or your camera phone and it can interpret those things as memories I mean there are people out here man they can know what's going on your mind because they can't reach your mind what they do is they reach your eyelids and your pupils David Blaine he's good at that s*** when you told that guy you know that the god of sanitation working he was saying let me see you like and he was put up a front like you trying to read it but he can't he's wait a minute and the guy began to stand there laughing you like ace of spade and the garbage man said son of a b**** and her walk though he's a ace of spades man so he don't read your mind he read the way your eyes and pupil reacts is it something new not really it's an old old technique but it's not known that much too many people that's how your muscles stretches and stuff and a person that's proficient in it can read your pretty damn good 👍 the CIA uses those technique when they interrogating people they don't have no magic book they just simply know how to read the eye pupils

[@LevendeStreg]

@@charlesneely Thank you for commenting. I'm so sorry to hear about your childhood trauma. At the moment Midjourney states noting about reading your camera and stuff like that. I'm sure it's doable. And I know that there are test with fMRI-scans and prompting. So that you think up a picture, and AI can recreate it.

[@biomuseum6645]

@@charlesneely which ai?

[@WatsonStreetPictures]

Would you know if you can specify focal lengths and exposure settings from a camera?

[@LevendeStreg]

Yes you most certainly can. But from what I’ve tried you need to be very specific about it. Like being specific about what camera you’re trying to emulate- what lense you want to use - and so on.

[@jenbraga9372]

Thanks Lavende

[@LevendeStreg]

You’re very welcome @Jen Braga. Glad you liked it🙏

[@CrhystyannDavidBernalChaves]

I love it!!!

[@LevendeStreg]

Glad to hear it @Crhystyann Bernal! So do I🙌

[@gfunk425]

With MJ ver 4 are the image resolution still below 300dpi? Curious on printability of generated content.

[@LevendeStreg]

I think it is, yes. But I just fix it in Photoshop and enlarge it with gigapixel!

[@hectorprx]

Gold , all the best

[@LevendeStreg]

Thank you kindly, @Hector Zayas!🎉

[@UtubeEric12345]

Just found this channel, ofc it gets subscribed :) But I can’t finthe cheatsheet on your webpage, am I doing something wrong?

[@LevendeStreg]

Thank you kindly. So glad you like my content. Go to this page - and find the video - and then there is a link in the corner: levendestreg.dk/en/tutorials/

[@ReidDesigns]

Holy. Plant seeds Batman! This is awesome! Thanks!

[@LevendeStreg]

Hahahahahaah LOL. That's really funny. You're welcome. Thanks for commenting🙌

[@nesrinebakhta2726]

Hello, first thank you somuchh actually, I am preparing my master's thesis in architecture about AI and architecture I need some information and scientific resources and documentation about some generative AI models and how they work, like their flowchart and their architecture, I have found some about DALL-E2, stable diffusion and Imagen and I need now infos about Midjourney AI, blue willow AI, Playground AI, Leonardo AI, and bing AI , I did not find any scientific resources about them, pls if you have any informations or scientific resource i will be thankful

[@LevendeStreg]

Hi there, and thank you for your question. Midjourney is Not Open source - neither is Leonardo. The other service, I don’t know if they have documentation out there available. Leonardo builds on SD, and so do most other services. They’ve just built their own UI around it. Your thesis sounds very interesting. Good luck with it all🙌

[@nesrinebakhta2726]

thank you so much for your reply@@LevendeStreg

[@Oxes]

awesome video, can you make a video on how to convert a photo into any drawing style? but for it to keep the same post and not be changed. thanks!

[@LevendeStreg]

Thank you kindly @Oxes 🙌 I think I would just use some of the Photoshop filters for that. Is there a reason to not use PS for it? What are your thought?

[@BettyMullins]

Which envelope do you need to use for midjourney to send you the seed number. I see several envelopes.

[@LevendeStreg]

I don't think it matters what envelope. You just need to press the envelope to create a DM for the bot (direct message). Then it'll send you a DM back with the seed number.

[@dr.ignacioglez.9677]

Muchas gracias 😎👍

[@LevendeStreg]

You’re very welcome! Thanks for watching🙌

[@N0melo]

Youre amazing 🎉👍

[@LevendeStreg]

Thank you kindly @NikeWilhelm - I really appreciate it!🙌

[@infographie]

Excellent

[@LevendeStreg]

Thank you kindly🙏

[@jameswatts3207]

Hi there does this seed thing still work???. I can not get any private message when I add the envelope reaction... literally nothing happens... just the reaction of an envelope is then attached the image gen "that's it"..... Your video is great, but this is why I have re watched and tried the seed trick but no success. You say "add reaction" then skip to saying "midjourney then send seed no./private message" but there must be something missing or they have disabled this now????.... making me a little crazy here, so hope you can help... many thanks

[@LevendeStreg]

I had the same problem in the beginning. So I wrote to one of the mods on Midjourney. He showed me that it sends you a private message - so you need to look for a message outside of where I was generating. Likely you need to click you “home” button in the top corner. There you should see the Midjourney bot as a DM (Direct Message). But you should ask one of the mods - because maybe you need to change something in your settings that allows the bot to DM you. Hope that helps🙌

[@resistancepublishing]

Nothing makes an “artist” better than finding ways to cheat as in using other artists creations to make it your own. But that is the future isn’t it.

[@LevendeStreg]

Thank you for commenting - I think most artist “cheat” and find the easier way to create visual content that looks good. And yes, I think this is the future. It’s also a sort of craftsmanship to be able to create with the help of AI.

[@resistancepublishing]

@@LevendeStreg thanks for responding to my comment. I hope that you didn’t think that my comment was about you. I was speaking in general terms. Even when I used the word “you” in my comment. I’m somewhat of a diehard purist fan of art. I was looking at the whole thing from the point of view that there will be people out there who will abuse the medium by saturating the market with AI art that they didn’t actually have to work at to create. New advances are always a double edge sword but I guess we’ll have to leave it up to fans to make the final decision in the end

[@LevendeStreg]

@@resistancepublishing than you for pointing that out 😅 - and I agree with you assessment. I think there will be an overflow of “bad” artworks out there. But it will only be of a shorter duration until copyright law and artists learn how to navigate this - and get better at using it.I hope anyway… 😉

[@resistancepublishing]

@@LevendeStreg yes we can hope but unfortunately artists are usually late when taking action when it comes to the business side of art. Mainly because they spend most of their time focusing on their craft. Like I said the customers and fans will decide in the end. There are two sides to art in 2022. The customers who just mindlessly devour every crap that’s created. And the traditional purists who support and highlight good art.

[@milsivonchip]

Levende streg - I må være danske? Fedt. Tak for tutorial, jeg vil da checke det hele ud.

[@LevendeStreg]

Ja da, vi er danske! Glad for du kan lide videoen, tak fordi du kiggede med.🙌

[@jopalange229]

hi, how do i find the cheet sheet? The link provided brings me you your videos.

[@LevendeStreg]

That’s correct. And for one video there is a button where you can download from! (Just under the video)

[@frykenmedia]

Great content. What is the main difference between remix and seed?

[@LevendeStreg]

Thank you kindly @FrykenMedia - the difference is that with seed you use the original noise to create a new picture. With remix you use the image (like img2img) to create on top of.

[@frykenmedia]

@@LevendeStreg Thanks. The reason I ask is that I would like to build my image in steps where I add elements for every new step (this is the procedure I understand the pros use for maximal control). However, this method requires a method for keeping the base before every new step and to this day I haven't really found a way. For example, I made an image of a roaring bear on top of a train (moving towards the "camera") to which I would like to add a dog on the train track in the foreground. To achieve this image in a single prompt is impossible I think. Another way would be to make a composite in Photoshop.

[@LevendeStreg]

@@frykenmedia yes Photoshop adding layers. But also try DALL-E outpainting. Og maybe Invoke outpainting. That would work, I think.

[@frykenmedia]

@@LevendeStreg I will try Dall-E. I believe this would be a good topic for a video, that is how to visualize an idea in a controlled way. But I understand that the creative unexpected interpretation is what most users want and according to some, the power (and nature) of Midjourney. Also, it is also difficult to prompt for the type of images that are very rare in the vast amount of images on the net. To build in layers or steps is necessary in those cases I guess. For example I am trying to make realistic versions of my favourite comic strips from Bizarro and The Far Side where it is not unusual that two snakes are having dinner while trash-talking a lizard that walks by. To ask Midjourney for this is not easy. Is Stable diffusion better for this? If yes, what is the best platform for SD in your view, both online (and likely paid) and locally?

[@LevendeStreg]

@@frykenmedia thank you for your question. I don’t know precisely what you are aiming for with the style and artworks, you want to create. Just remember that copyright wise you might get into trouble - and it is not advisable to attempt to copy other artists’ material. Yes, Stable Diffusion let’s you train models. So it’s much easier to do with SD.

[@cryptidsNstuff]

Good job.

[@LevendeStreg]

Thank you @PromptCrafter! I really appreciate it🙌

[@jrvideoca]

I went to cheatsheet url, its 4 video tutorials? Is that right?

[@LevendeStreg]

Yup. There’s a link button to the cheatsheet!

[@kimberlyjones4112]

Hi, my name is Lee. I'm new to the whole mid Journey scene. I'm having an issue trying to produce a seed I copied the address linked to the image that I wanted to work with I clicked on the Emoji. Initially it did not show the envelope so I typed in envelope manually, I clicked on it and it actually sent me a message under the image. I checked to see if it produced another image with the seal number above it but it did not. Although it did produced an envelope underneath the image which I clicked on but it did nothing. What am I doing wrong? Please help!

[@LevendeStreg]

Great question. You did the exact right thing. But you need to allow the bot to send you information. 1. So you click the channel that you added the bot to. Make sure that DMs are allowed. 2. Click the settings - and under “roles” and then “permissions” make sure the bot can send messages.

[@LevendeStreg]

And Hi Lee. Great to meet you 🥳

[@1-o-f-1]

Can you explain how to add the seed back to the prompt to be remixed.

[@LevendeStreg]

Sure - you just copy and paste. I show in the upcoming video!

[@1-o-f-1]

@@LevendeStreg Thank you

[@mrxXxkink]

I'm trying to upload my logo and just change it's background but it keeps changing my logo as well. Can you help me?

[@LevendeStreg]

Thank you for your comment. It you’re using DALL-E you need to erase the background of your image. Otherwise the engine thinks it should hochlange everything.

[@gumvue]

hi, I have a script for a movie, and I love to build a comics storyboard and generate all my book with this AI , is that possible,, how bad it can be ?

[@LevendeStreg]

Thank You for you question , @gumvue. I thinks it valid to say yes. I’ll take it up in my next video. I have to say that you probably need to use a Mix of Stable Diffusion, MJ, DALL-E and Photoshop. But sure, you can make that work!

[@zhurst]

It is technically training with DreamBooth. It’s not training with textual inversion. Regardless of semantics, it’s an amazing form of fine-tuning existing AI models. Nice video!

[@LevendeStreg]

You’re totally right @@Zach Hurst. Thank you for pointing that out. I’m trying out stable diffusion at the moment, but somehow I’m not as impressed with that!🙌

[@zhurst]

@@LevendeStreg Emad has 3 larger models in training. SD 1.4/1.5 is just a tease. What’s coming will be absolutely mind blowing.

[@LevendeStreg]

@@zhurst Thanks for commenting 🙌 - Are we talking Stable Diffusion (Stability AI) - I'm trying it out now. I'm still not that impressed with it. But I still have to try out some more to make a video on it. At the moment I'm not getting so good result with it when it comes to comic book style.

[@zhurst]

@@LevendeStreg Yes, Stable Diffusion. MidJourney’s latest updates really upped the game but new models from SD are coming too. There are also MidJourney models underpinned by Stable Diffusion. Every model has a unique latent space so prompts don’t translate. The difficulty right now is learning to prompt every engine so a number of folks are sticking to learning one or two. Large scale language models are going to solve this imho, we just need to wait for GPT-4 or maybe 5 but I’m betting on 4.

[@LevendeStreg]

@@zhurst I can hear that I need to look deeper into Stable Diffusion. I'm running it through DreamStudio at the moment and trying it out. But I'm not getting as good results with that as I am with Midjourney. Thanks for sharing your knowledge with me, @Zach Hurst!🙏

[@Panda]

Very col Levende :)

[@LevendeStreg]

Thank you kindly 🙌

[@mistycloud4455]

A.G.I Will be man's last Invention.

[@LevendeStreg]

Hopefully we'll have many more cool things to invent. Lots of good can come of AI. And yes, lots of bad things can come of AI. Let's hope and pray for the good.

[@MrMadact]

Brilliant, incredibly useful, and... a little twinge of pain whenever you say "so why aren't everybody else using this?" haha. Maybe it's an American dialect thing though?

[@LevendeStreg]

Thank you @MrMadact for your kind comment☺Please explain the dialect thing? I'm not American (but I do have a large American family on the west coast of US). I'm actually Danish (like the vikings and Lego)... So am I pronouncing something oddly? Would love to know so I can correct it🙌

[@MrMadact]

@@LevendeStreg ah, that would explain it - I presumed American from the accent. It's the plural matching of "is" vs. "are" - you could say "why aren't others using it" (or "other people"), or you could say "why isn't everybody using it" (or "everyone") but "aren't" doesn't quite match with "everybody". One of those silly language things where the "every" implies that it's a plural, but it actually goes with the other part of the word. "Each" works similarly, so you'd say "each tree is bare" or "each of the trees is bare" but not "each of the trees are bare". I think I found it jarring because apart from that you sound like a native English speaker.

[@LevendeStreg]

@@MrMadact Ahhh, that makes sense. Actually my first (intuition thing) was to say "why isn't everybody else using it". I had to do a small number of retakes on the sentence... hahaha.. Because I kept getting it wrong - turns out I kept getting it right - and picked the wrong version to go in the video. That's hilarious. Thank you for calling my attention to it😂

[@sarabethmattis6099]

For some reason, I cannot get a modern trailer in a trailer park with a deck and hot tub.

[@LevendeStreg]

Hahaha… sounds cool though.

[@user-ut4hj7kc1t]

As far as i understand same seed does not work with v4, is not it?

[@LevendeStreg]

That’s correct, so far it only works in v3! 🙌

[@perfectone8185]

plz where is the link to free PDF ? I'm lost couldn't find it

[@LevendeStreg]

Under This link - just see under the video there’s a button : levendestreg.dk/en/tutorials/

[@perfectone8185]

@@LevendeStreg thank you so much ^^

[@LevendeStreg]

@@perfectone8185 you’re very welcome!

[@Kyus2001]

Stable Diffusion and Motion capture! Plask ai

[@LevendeStreg]

Thank you for commenting, @kyus2001. I would love to learn more about motion capture. So maybe in a future episode.

[@theXaint]

2:50 cute blond eyes, metal viking Ham-sword - Can you do this prompt for me? :-)

[@LevendeStreg]

Hahahaha... I think it need to be a little more detailed than that for the prompt to work 😬

[@theXaint]

@@LevendeStreg Haha! I've find it funny that you typed cute blond eyes for a badass Viking! It would be fun to see a ham-sworded metal viking with cute blond eyes in your upcoming video. Just for fun, if you feel like playing with it, expand on the prompt! See what pops up. I'm just getting into this stuff, like many others, so your videos are great!

[@LevendeStreg]

@@theXaint hahaha. 🙏 The cute eyes are weird - I know. You could just as well write detailed eyes. As long as you remember to mention the eyes. MJ messes up the eyes, if you don’t remember to mention them in the prompt…🙌

[@theXaint]

@@LevendeStreg I would have imagined midjourney dropping images of Hemsworth with those big puppy eyes, like one from a dog food advertisement. 🥺 LMAO!

[@piggywahwah]

how u get the seed? i dont see it at all :(

[@piggywahwah]

ok, im being lazy, quick google search and its the envelope emoji ;). Thanks

[@LevendeStreg]

@@piggywahwah yup. It’s kinda DM to the bot!

[@LevendeStreg]

Hey by the way - you can join my discord, where you can ask people about stuff like that(remember to agree to rules - otherwise you can’t post in there): discord.gg/bu5HQEYd

[@mariosuze]

as of today, the reaction to your own photo does not work :-(

[@LevendeStreg]

Oh, I’ll check that out and see what’s happening. Thanks for pointing my attention to it 🙏

[@MentalPistol]

Creative woman + House looks clean = possibly wife material

[@LevendeStreg]

Hahaha… it’s my Office - I’m not doing the cleaning - I’m the boss woman.💪

[@sarfarazusm]

you beauty

[@LouisGedo]

👋

[@dmaxst]

wow nice body language 🤠

[@LevendeStreg]

Hahaha thanks 🙏

[@SaberBE]

i quit watching after the 3rd reminder to hit the subscribe button

[@LevendeStreg]

Thank you for your comment and for watching. Have a great Christmas!

[@KenjutsuDojo]

7:51 Jump here for the "real" content

[@LevendeStreg]

Thank you kindly for watching and commenting🙌

[@KugleeKuglee]

Midjourney a lot of times is just not care about the right command (like full shot, full body etc...).

[@LevendeStreg]

True, but I actually have an upcoming video on how to cheat a bit and get it to follow your directions😜

[@KugleeKuglee]

@@LevendeStreg cool!

[@KugleeKuglee]

@@LevendeStreg I tried Leonardo AI and looks like we have much more neat controls and the interface more friendly with lots of help.

[@LevendeStreg]

@@KugleeKuglee I’m doping a video on Leonardo soon🙌

[@KugleeKuglee]

@@LevendeStreg ❤

[@hungry9003]

8:05 seed number

[@charlesneely]

I don't know about all that but I do know this you're very beautiful you're very pretty and you are the kind of person I would like to get to know so unfortunately too bad I can't come to nowhere else we know wherever you're at LOL but I just want to say that

[@LevendeStreg]

Hahaha... thanks @charles neely.

[@TheKenwahl]

more vikings, great

[@LevendeStreg]

Haha. Thank you kindly @TheKenwahl😊

[@biomuseum6645]

Those Vikings simply can’t be ai made, I searched in google and no MJ comic style resembling this one They still have the warmth of human art, not the coldness and cheapness of ai images

[@LevendeStreg]

I assure you, they are. I can upload the video and show you my feed in Midjourney. I corrected slightly with his hair. But otherwise made by me and Midjourney🤗

[@biomuseum6645]

@@LevendeStreg please upload the video 👍🏽

[@Tsevries]

google doesn't even begin to cover what a search of the actual new-images community gallery would show. Most google results for midjourney are wildly out of date, as the AI gets better every few weeks. I've been trying variations of her prompts and they work great. Not everyone knows how to write such good prompts so it's natural that a lot of the images won't be like this one.

[@biomuseum6645]

@@Tsevries this is beyond sad, I swear I want to f k1ll the devs of this abomination

[@LevendeStreg]

@@Tsevries thank you 🙏 and I would love to see how your prompts work out. And you’re right - so far as I know you can’t Google prompts and AI images…

[@sergiosergiou6551]

Hi there! I got a challenge for you, create a way to prompt mythical creatures or any creatures,animals fighting each other

[@LevendeStreg]

Hi there @Sergio Sergiou. Thank you for your request🙌. That's quite a challenge. But right now I'm more focused on creating styles and comic characters - and sort of tweaking and training Midjourney and Stable Diffusion. But I'll definitely keep your request in mind.

[@MrLaura34]

I like midjourney, it's not for free! if you pay x month. I do not have anything.

[@LevendeStreg]

I love Midjourney too. But I use a lot of AI platforms!🙌

[@MrLaura34]

@@LevendeStreg And I don't have much left on the platform. I already know it. You can't access my cell phone number and email if they blocked me, it's theirs. all wrong.

[@Phil_Langone_FilmArt]

As an artist AI is one of the more depressing developments in art history. Obviously it's a wave that can't be held back. I'm sure the goal for most artists is to figure out how to control this for their own work, but even that, the time it takes to learn, the time it takes to write prompts, it's just a depressing, boring process.

[@LevendeStreg]

I tend to agree with you. But there are also a lot of potential in using AI. I still prefer to illustrate, draw and visualize🙌

[@Phil_Langone_FilmArt]

@@LevendeStreg there is a lot of potential. What that is while not wrecking people’s careers I’m not sure yet. In LA, within the concept art community and larger film art community there’s a lot of worry and meetings happening. People that have built overhead, with homes, families, bills are seeing all the horrible things that can go wrong for them. It’s already happening with producers cutting out concept artists and paying low paid PAs to generate images.

[@LevendeStreg]

I agree with you, @Philhockey22. It's concerning that currently, only major studios like Disney, Pixar, and Dreamworks are utilizing AI to create 2D characters in 3D spaces, as it requires significant computing power and advanced graphics cards. This technology is out of reach for individual animators and illustrators due to the immense resources needed. However, I believe that as AI becomes more affordable and user-friendly, many creatives will integrate it into their workflow. For now, AI doesn't seem practical for smaller projects, as it doesn't yet offer time or cost savings in those contexts.

[@jeremyp2164]

the fact you have a cintiq wacom sleeve is fucking hilarious and you're saying what you want in your style even though you gave the program prolific artists to reference. I hope you know this will just be used for recreational use and be used for pros in the industry. their is no way you can get a art job working in a studio without actually knowing how to do art right? Listen if you enjoy using midjourney then that's cool but dont' think you will be ever work in a studio or get hired without actually knowing how to do or make are by yourself.

[@LevendeStreg]

Thank you for your comment. If you watch more of my stuff you can see that I maybe use AI for 2% of the work I do. The other 98% is drawing and animating. You can check out some of my illustrations on instagram. And my company (that I am founder and creative director of) creates animation videos and digital comics for other companies. You can check out our website😊

[@LevendeStreg]

Oh and I do a lot of live drawings / graphic recordings - you know where you draw live and analogue while people are watching. No possibility of cheating in any way… that’s why our company is named Levende Streg. It means lines that are brought to life - or lively lines.

[@mio3456789]

Robbbers

[@joshuaborner]

I dont think its okay to use the style of living artists without their permission... This can be so hurtful and an actual danger for their livelyhood. They spent years to master their own style and achieved something that is very unique to them. I hope copyright laws will catch up with ai and prohibit the use of living artists names in prompts (if they dont consent)

[@LevendeStreg]

Thank you kindly for your comment. I understand where you’re coming from. For me it’s “easier” to prompt a style name thank to write a description on that style. But it’s more than doable to not name names. Also I advice to use more than one reference in a prompt. I am collecting my own drawings and will train AI in that style. I’ll do a video on it. But until I have that trained model, I probably will use artists as reference. I assume that you think the AI prompts and the illustrations created with AI will devaluate the artists’s work. But I don’t think it will in any way. Their artwork will still be much more valuable. Also it’s important to note that under no circumstances is it allowed to sell drawings under the pretense that it is made by the referenced artist.

[@joshuaborner]

​@@LevendeStreg What I mean is that their style might become very common due to ai dumping. So basically they loose their uniqueness. But in the end AI is going to take everyones job anyway very soon... Just kinda sucks that it seems to start with artists ( Who mostly struggle even without ai competition especially now after the pandemic )

[@LevendeStreg]

Hmm I understand your point. You should check Jazza video on AI. he just did an excellent one on it showing how artists can really benefit from using AI on top of their own work. I’m going to try out something similar. I think artists can greatly benefit from AI…

[@LevendeStreg]

This is the video I’m referring to: kzbin.info/www/bejne/ioDJdn2vjLOmkLM

[@underdogunchained]

Too much talk and not enough practical examples to help the viewer understand the process.

[@LevendeStreg]

Sorry that you think so, but this is a tutorial for the setup. Next episode’s a tour of the dashboard and the different parameters and how to do prompts the best way.

[@pawnoir]

12min51 to say you need to use --seed

[@LevendeStreg]

Hahaha @pawnoir - thanks for watching and commenting. Glad you got the gist of it (through all the information about cropping, 3D, Clip Interrorgator, Stable Diffusion, img2img, cheat sheet and so on)...

[@pawnoir]

@@LevendeStreg I couldn't tell which one was 'the secret', i picked up the less known one.

[@liquidmind]

@@pawnoir she is making a tutorial for ppl who need it, some others already know how to do this. and yes many content creators even make 1 minute videos JUST TO SAY HI, or JUST TO SAY "hey, v4 is out" and they get their views alright....... so..... ????

[@wejd733]

You're talking so much get to the point i feel every video you do is a trailer for upcoming video

[@LevendeStreg]

Thank you for your comment. I am learning all the time as I go. The video you made this comment for is a couple of weeks old and already I’ve learned so much since making that🙌

[@MikeKleinsteuber]

Well it ain't quite as simple as this. Using seed is very random and most of the time it doesn't help at all at creating a similar look (it may of course for comic characters like yours). Also your continual insistence on subscribing is really intrusive, amateurish and boring. Just back off trying to push your channel...

[@LevendeStreg]

Thank you for you comment, Mike. I’m just getting started with my channel (even though I’ve had it for many years). But I only just this year started to make KZbin content. - so I learn as I go.

[@MikeKleinsteuber]

@@LevendeStreg Seeds get randomised over time so this idea of using the same seed to get similar results needs to be done over a very short space of time....

[@leaphardotnet]

How to train midjourney ? Well you can't! Please don't clickbate

[@LevendeStreg]

I'm sorry you see it as clickbait, @leaphardotnet. But for people who are new to Midjourney, they don't know what it's called to use --seed or --sameseed. I know you cannot do textual inversion in Midjourney yet. And a lot of people don't know what that's called either. So when they search for a solution to their problem they're gonna search for something like "training" or "how to train". And here I offer a solution or workaround to the problem. So that's why I have that title.

[@guruware8612]

why are people lately talking a lot with their hands, there is no relation of movements with the words spoken - should this make things more exciting ? stopped after 20 sec - disgusting it is

[@LevendeStreg]

Well, I don’t know why other people are gesticulating with their hands when they talk. I’ve always done it. Don’t think it’s a trend. But something that comes natural to some people. I’m sorry that it triggers you and bothers you so much as to call it appalling and needing to comment on it, though…