The Lady in Black | Watch Me Prompt Live in Midjourney

preview_player
Показать описание
Folder: 2023-05-20 – Watch Me Prompt "Lady in Black"



🎬 The "Watch Me Prompt" Series 🎬

📺 Recommended Related Videos & Playlists 📺



IMPORTANT: This video is intentionally non-scripted. It is a more or less raw recording of an improvised prompting session. This is content that I create for my regular subscribers who enjoy this type of content. If I speak too slowly for you or you think that I should get to the point, then with all due respect, you clearly don't understand the point of these videos. Feel free to rant in the comments, but I'm just going to ignore you anyway :)

I realize I haven't been publishing as much content recently and it's mostly because I've been working my butt off on my course content.

It's really tough juggling between social media, creating YouTube content, and also structuring, scripting, recording, and editing a full course.

That being said, I don't want to leave you guys hanging which is why I at least want to release another one of my "What Me Prompt" videos.

I actually think you'll find this one particularly interesting because in this video I demonstrate a prompting technique that I call "Layered Multiprompts".

It's a really useful approach to creating images in Midjourney because it almost feels like you're painting on canvas.

The technique is so useful and pragmatic that I actually need to publish a proper video just about this technique but I'll have to save it for another day.

Credit goes out to @Danner, one of the trusty Moderators in our Tokenized AI Discord who also happens to be a Guide in the official Midjourney Discord.

He's the one who originally inspired me to formalize this method a little more, so shout out to him for this excellent idea.

Enjoy the video!



⏰ Timestamps ⏰
00:00 Watch Me Prompt "Lady in Black"
01:16 Creating the Base Layer
03:03 Adding a 2nd Layer
05:10 Adding a 3rd Layer
06:21 Retouching the Look
07:49 Fixing the Colors
09:10 Increasing Coherence with Mirrored Multiprompts
10:53 Permutations



#midjourney #midjourneyai #midjourneyart #aiart
Рекомендации по теме
Комментарии
Автор

Very nice, I followed your videos and man I have to say: you should give some courses. Very valuable content!

cvdheyden
Автор

Laughed at and replayed your "male/female" commentary at 10:45 -- So true, lol!!

turkeybuttsammich
Автор

It is interesting that you referred to "wide angle shot", "glamour shot", "Photoshoot", and Midjourney prioritise this over the style reference which was in the begining of the prompt.

teolita
Автор

Thanks, Chris! It's always reassuring to see that I'm not the only one who struggles with getting MJ to give me what I actually want. I struggled for days trying to get an image of a river or stream or creek to run left to right rather than front to back, and I eventually gave up. I wonder if your project would have worked better by creating the background in one image (as you did at the start) and the model in another image, and then combining in Photoshop. It seems the more you ask of MJ, the more it begins mixing things up between foreground and background. Anyway, thanks for sharing your struggles! (Oh, by the way, I don't think "canyonous" is a word, although I know what you mean -- and it seems MJ understood, too!)

harrygoldhagen
Автор

Glad to see you explored this! 😊 you should try this with adjusting the weights now. 😉

PhillipDanner
Автор

Great walk thought the steps! However - I noticed that you did not get to the point where we really get an image which you chose for you thumbprint…..The video ended with unfinished goal “a woman wearing a black business suit” . But we see you did it….I wonder how many more reruns did you need…? Otherwise - tis is phantasmic! Comprehensive for someone who wants to get good at this.Thank you Chris!

StyleViewStudio
Автор

Thanks for another interesting technique. I imagine that doubling up the prompt descriptions provides a similar result to using an image prompt, but this method gives you the ability to include separate image weights, which unfortunately you can't do with image prompts... yet. ie the ability to allocate different image weights to individual images within a prompt that includes multiple image URL's.

morpheus
Автор

PS. As a Fashion designer of many years - please know that there are hand drawings in fashion which get closer to photography - some are watercolor based others made in pastel or ink. Emphasis may be on the flair of the coat, shirt etc - glamour is not always what we need. All you need to do is scout any good fashion magazine and you will know what I am saying.

StyleViewStudio
Автор

That can also be a good method to add multiple characters in one scene.

milestrombley
Автор

Thank you! I honestly thought you were going to try weightage for each segment.
Also, it may work for you, but after you've typed a prompt and it gets started, you can press the up arrow and the entire thing will be copy/pasted. Saw that trick somewhere. It helps, a lot!

AG_before
Автор

The black suit doesn't really goes well with the overall colors Midjourney chose. The colors MJ kept putting on the suit were better than black. 😅

teolita
Автор

Thanks, Christian:

That trick of emphasizing and re-emphasizing the thing you want to have stand out in a multi-layer format seems to definitely work. I'm working on a graphic novel that includes a female ex-Navy pilot who has blue hair and tattoos which she had during her time in the Navy. I wanted her blue hair to stand out in pictures from her Navy days while in uniform. Nothing I tried worked until I simply tried mentioning in a multi-layered fashion that her blue hair and tattoos should figure prominently in her depictions:

/imagine PathtoURLjpg:: she is clothed as a U.S. Navy pilot:: her blue hair and neck tattoos are exposed outside of her uniform:: her blue hair and tattoos are part of what made her a popular Navy pilot:: members of the military were proud to have her display her blue hair and tattoos:: --v 5 --s 750

I now have more blue hair and tats than you can shake a stick at. An ungodly amount. They're in every picture now.

If only I can get my Underdog cartoon character that I am working on to have a blue cape (like he is supposed to have) rather than a red one that matches his red bodysuit (it's not supposed to match) then I can die a happy man.

timothypruitt
Автор

Good stuff, thank you! How would you prompt for a non-centred portrait shot, more towards golden ratio?

kurkista
Автор

Love this, but curious why you preferred v5 over v5.1?

turkeybuttsammich
Автор

Curious at 12:43, what if you would have clicked v1 or v3 under remix, and edited the prompt by emphasizing "black pantsuit"?

turkeybuttsammich
Автор

Thanks so much Christian! Do you have a public Discord server we can join?

michellebracken
Автор

Maybe I'm missing something but the woman with sunglasses in the video still is another prompt, right?

CaseFixed