I'm not sure if I explained this in the best way possible - let me know if I missed anything!

It's the best seed explanation that I've seen. Future TECH Pilot is a favorite channel for quality information, hands down!

It appears that "Gutai Group" might have encouraged multiple faces in the results. In its absence, there was only one portrait in each image. That suggests to me that the powerful part of "Gutai Group" is the word "Group". Might want to test "Gutai" on its own.

This is the best explantion on Seeds and Premutations I have seen. thanks!

This is 100% how to prompt build from scratch or playing with an existing prompt I call it "Prompt Dissection" - I find myself doing this constantly as we get a new version or large update. Nicely explained.

I love your lessons. the best explanations of all the nuances. Thank you // обожаю ваши уроки. самые лучши объяснения всех нюансов. спасибо

Priceless tips as always Ai MJ brotha 🎉

❤❤ thanks for efforts you did for us❤❤

nice tip. the complicated prompts ive been inputting, it would be nice to see what phrases dont really matter.

Your efforts are gold!

You're videos are always amazing, thank you.

The shorten command still help. If i place gutai as the first word, it gers 0.77 tokens. At the middle, it gets 0 23 tokens. I'm not sure this is updated to V6, but it seems to reflect on the images in my tests. Great breakdown 👌👌

Completely excellent tutorial. Awesome. Thanks.

Thank you for the advice...

Very helpful. How do you find these gold nuggets is beyond me. Keep them coming!

Amazing video! Thank you so much for this. You mentioned adding "anime hair" to the prompt didn't do anything, but it did, well, not so much to the hair lol. The woman in the photo has more of an Asian ethnicity in a few of the images due to Anime having strong ties to Asian culture. That was really cool to see.

Killer FTP! Thanks!

Thank you, Nolan, for finally proofing that all those "use seed number for character consistency" are wrong 😅

„Speeding“ and „racing“ included more cars in the final image opposed to „hurrying“ - I thought that might be helpful ;)

hi Nolan thanks for the excellent video. How do you get consistently when you hit on a style you like?

In v 6 without zoom, I find the cropping way too tight. Any descriptors to get the frame zoomed back some? Odd MJ hand added zoom and pan yet. They could have just used the code from 5.2?

Thanks for this great video! Is --sameseed still a thing too? I've never quite understood how and when to use it.

So it sounds like seed numbers just provide a similar style to the image it corresponds with once assigned. There are a lot of videos claiming that seed numbers help generate the same characters which is just not true. Once we figure that issue out it will be a game changer.

This is genius. I earn a bit of money with Midjourney, but maybe I should give it to you 😅.

Okay, so by making the SEED -- assigned and constant, no longer variable we are able to determine the strength of our other words/variables within the stacked equation/prompt. That's the argument, right? I hope I'm warm with this, thanks for explaining. :)

thanks

Great idea, there's also the "shorten" function to inform how much the token is affecting your prompt; but seeing it visually feels more controllable

i still cant find solution to make real consistent persons with same faces on different locations, even when i use seed or --sref its still make different faces, do someone have solution ?

If i was doing real life like photos of people smiling and use them as a hero section image, would these look real enough to use? Just wondering also if you found a fake life like model how can you keep this person/character for other scenarios? Thanks for the content it’s awesome 👏🏼

Who else was excited to go use permutation on relax mode? (>_

'cartoon anime hair' turned the model Japanese ...

Hello guys...I hope a fellow graphic designer can help me out please? I am currently creating a fictional book with images that were created pre-version 5...I am having great difficulty in re-creating the same style in version 6 Alfa (Jamie Hewlett: Artist) can anyone advise me on how to stabalise the bot with a generic prompt that i can use for all new images please?

I need help. I specialize in mermaids, sea sirens, water nymphs; no they are not the same. Anyway, when you include the word "underwater" in any prompt, 99 out of 100 times MidJourney will include the 'surface' with 'sunlight' or 'light rays' in every result. Even when specifying not to include sunlite in the negative prompt. So, here is my challenge: please give me a reliable, consistant prompt that when I write... "mermaid swimming in deep ocean, on a coral reef" will not have any surface lighting effect. I've tried a thousand times but unsuccessfully. The only way I come close is including in the prompt that it is an "overhead" view, which limits me. I want a stragight ahead shot of a mermaid swimming just above a coral reef with no sunlight or sunrays.... Can somebody help me? Thank You.

👋

..

they blocked me for no reason clarified. ditch mj as soon as you find alternative

❤❤ thanks for efforts you did for us❤❤