NEW Midjourney UPDATE is a lot of FUN
Vložit
- čas přidán 5. 06. 2024
- Midjourney Beginner Course ➡️ ftpai.thinkific.com/courses/m...
Midjourney updates brings a new feature before Version 6.1 - it's called SREF Random and it generates your prompt in a random style each time you run the bot. Take a look at the breakdown in this video
Podcast Link ➡️ / @futuretechpodcast
► FREE PDFS
Prompt Glossary V6 ➡️ ftp.live/V6Glossary
2-4-2 V6 ➡️ ftp.live/242
200 Aesthetics ➡️ ftp.live/200Styles
Midjourney Cheat Sheet ➡️ ftp.live/CheatSheet
► SOCIAL
Discord - / discord
Instagram: / futuretechpilot
Timestamps:
0:00 Intro
0:15 How to use it on Discord
2:12 How to use it on Website
3:13 /Describe on Discord
4:05 /Describe on Website
5:10 Extending Style on Website
6:05 My Favourite Way to use this Feature
Dude love the Midjourney vids. Thanks for putting me onto this
And thank you for the comment!
Great video! Thanks, this will be useful for future content ✌ - subscribed!
I appreciate the support!
Thanks! I'm liking how I'm getting to see some styles I've never used.
For sure, who knows what's behind the next corner haha !
Good results! I'll have to look into that combo of chaos + style. Thanks for the update and tips and tricks as always! :)
And thank you for the feedback! Enjoy :)
this is really amazing. thank you.
And thank you for the support!
Wow the alpha site helps alot thanks for the tips man. Sref random is so fun and really helps when you get burnt out om ideas.
I completely agree!!
Woah, I checked out for a few months and this was an amazing succinct catch up!
haha cheers pal!
Using MJ is becoming more and more like computer programming. I don’t use it as much anymore and I forget more of the commands than I remember.
😂 I know exactly what you mean. It's difficult to keep all the options in the top of your mind
It does feel good to get back to some fun, creative stuff, being surprised, especially after hours spent on complicated, boring images for clients. Maybe I'll suggest Random Batman, maybe they like it
haha yeah it's a good shakeup for sure
Thanks!
and thank you!!
I like this a lot
enjoy!
Thank you for another good material :) MJ is still doing its best to rule image generators!
I agree!
Great per usual! Keep em coming Sir!!
Cheers pal!
notifications from you is my favourite thing
and I really appreciate the support!
Thanks for this amazing video! Do you have a link to the random batman image you created?
s.mj.run/odVO72GiE_g - that should work!
very cool new feature
💯
I told you a long time ago in v5.2 style random is the best feature!
haha yeah it's the best slot machine ever 😂
Nice to see another vinyl collector. Blondie and Eagles both good choices.
haha my Mom was cleaning out her basement and getting rid of old records, so I saved a bunch that looked cool!
Loving these. Spin the random SREF and stand back!
😂 Midjourney is the best
Interesting….thanks!
Cheers!
Hi! Thanks for this. You did forget one thing --sw .5 to --sw 3 can also be used.
Maybe I'm a little confused ... I thought --sw was 0 - 1000 ?
@@FutureTechPilot Oh I am a dummy, I got it confused with something else that was .5 to 3. Now I doubt my sanity.
@@RichardSekmistrz 😂 there's a lot to remember in Midjourney
@@RichardSekmistrz that's --iw :)
@@latentspace_fi Thank you! I thought I lost my marbles ;)
On the website, the "+" besides the prompt field will show you every image that has been *uploaded* to the site, not any images used as image-blend/style/character refs. If you drag an image from your archive into the prompt field, it won't appear in this image area. On a related note, I am not aware of any way on the web interface right now to describe an image without downloading it and then uploading (or pasting it in the prompt field, if the image is on the clipboard) into this area so you can access the Describe feature. It would be nice if describe was available directly for any image.
Yah that's more accurate, sorry! I think I meant it from the Discord perspective, where any image link you use will show up there on the Website - but you're right, if you just drag and drop into the prompt, they won't be saved.
And no, it doesn't look like there is a way to access /describe unless it appears in that '+' grid - but I also didn't know about describe being there so maybe I'm missing something else haha
1. Thank you.
2. Feel better, man. :)
😂 I try not to mention when I get sick but I really couldn't hide it in my voice lol
What comes first: your Top 50 images video or gpt5?
lmao you're right ... we've been due for GPT 5 for a while
I'm really shocked Midjourney gave you 4 different artist names in describing the batman picture and then 2 different ones again on the website. I'm really surprised it didn't stick with one artist. Are all 6 artists truly indistinguishable to Midjourney?
That's a really interesting observation and I guess I didn't give it much thought ... so I just ran all 4 prompts and maybe the difference can be thought of like this ...
Names of Artists are just tokens like every other word. And each word/token has a different meaning to the bot, each word triggers a different outcome. But there are only so many ways to describe a certain thing - so even though the different artists clearly result in a different looking images, their names can all be described in a similar way.
Does that make sense? Maybe it doesn't. I'm just trying to say that most artists are distinguishable by Midjourney, but some artists can trigger the 'same' look (for all intents and purposes)
Ah yes, another feature 90% of users won't use anyways! Thanks MidJourney! 😭😭😭
😂 aint that the truth
I love how you got into the website features. i feel like people dont use as much because they feel it lacks features, but it really doeasnt, i love the website, oh and by the way, your triple, triple, from your other video...AMAZING results!!!
Yesss, big shoutout to the 3x3 - so easy to do on the website lol it's actually a lot of fun
Thanks for the tutorial
I'm happy to help!
Good morning from Spain, great job you are doing to disseminate Mid Journey, I have a question for you. Until now, I have put all the works that I did not like in HIDE, so as not to view them. I work with the alpha version and my question is how we can definitively delete the files that we are not interested in. On the other hand, in the same tab comes the publish or not publish tab, which means that if public it automatically goes to public in the explore? section. Thank you very much for your attention, this last one that you have published is fantastic and it is a little scary with the ease that all types of works can be replicated, when I improve hands and feet and small faces this is going to be fabulous. Good day
So in your first batman trial where the images have a yellow tint, would there be a way to make wonder woman in that style but, say, with a blue tint?
hmm ... it's a good question. Theoretically you should be able to but it's a little tricky. I tried 'blue wonder woman --no yellow --sref s.mj.run/odVO72GiE_g' and it kinda worked but not really
actually, try lowering the --sw 'blue wonder woman --no yellow --sref s.mj.run/odVO72GiE_g --sw 75' .... or even try --sw 50
I wonder if adding an all-blue image as cref like a tip you had in an earlier video would work.
@@jasonp9508 Maybe! Good idea
good tip FTP 👍🏻
Cheers!
If you just put "batman" and nothing else you'd pretty much get four random things anyway, right?
I definitely should have shown an example of this. Sorry about that. Just putting 'batman' alone would have created a very generic Batman look - and it would be the same for everyone. But if we both put 'batman --sref random' - there is no way we would see the same looking generation
how do you get to have the same Kelly character? I am pretty sure you created a video where you can just type the name Kelly and somehow MJ knows who you are talking about, but I missed it.
Can you point me to the source, please?
Yah sure! This first video definitely goes over how to create shortcuts - czcams.com/video/v_Ni-vr6NQ8/video.html -
but you can also check out this other one for some more tips - czcams.com/video/zgeI_ffCudY/video.html -
hope that helps!
What style are these ?
What do ya mean?
@@FutureTechPilot these batman and Wonder Woman art style in this video , what do I type in prompts to make these type of art ?
Your videos are amazing. One on upscaling images would be really useful to your community please.
Okay thanks for the feedback! Specific requests like that are super helpful, I'll start taking notes now!
I just wrote 701 words - I think I'll have enough to say for a video for sure
I had some interesting results with this.
Having a lot of fun with : comic book art --sref random
Your polaroid idea is awesome! Thanks!
dang, thanks for the suggestion! I'm going to try comic book art now haha
Man... you are still the best!!!
I appreciate the support!
👋
🤝
So the Midjourney website isn't for everyone yet, is it?
not everyone, but I think he mentioned opening it soon to people who have made 100 images, so basically everyone. And then you'll be able to sign up for MJ without Discord at all
@@FutureTechPilot Man, I've got like 1000 images now. At least.
@@WilliamAlanWebb go to - alpha.midjourney.com - it should work for you
The slot machine pixel vomit
😂
MJ is now allowing people who generated 100 images onto Alpha.. so by this summer it will be open to everyone , it looks like. Alpha has certainly helped me fall back in love with MJ.
Yah it's super fast haha pretty fun to use
Is mid journey free
nope!
Listen: I like you and your channel, but I feel like 90% of the people who use Midjourney are Storytellers, and your examples never cater to us. Which is to say, I have never (in my 11,200+ images) used a prompt as simple as, "Batman", or "Wonder Woman" or "a penguin" or "a Polaroid of a model". And of course I understand: you're attempting to "simplify" your prompt as much as possible to highlight what the featured parameter can do. But... that would be like, had you a fashion or clothing try-on channel, only putting your outfits on a non-anthropomorphic, stick figure mannequin each video, because you don't want the body type to distract from the clothing (ie, curvy or big girls would find your channel pointless). And I get it, maybe you're NOT a Storyteller. But as a businessman and educator, you need to reconcile the fact that there's a difference between "a penguin"... and "a low-angle shoot of a penguin wearing clothing, hiding behind a dumpster in the meatpacking district of NYC, injured, with blood on its face and tail". If I ever used "Batman" as a prompt, it would probably be, "Batman, sitting wide-legged in a geasy diner, eating a sandwich; he seems oblivious or uninterested in the protestors who have gather outside, on the street, angry and screaming their lungs out." Yes, I know you sometimes utilize more complex prompts, but its less and less frequent. But this is just a reminder that MOST ALL of the people watching your videos are Storytellers (and that includes those with business/advert/marketing backgrounds); we're all trying to tell a story, so your one-word prompts are often a turn-off to people like me. SOLUTION: for these parameter videos, please try and make at least ONE (1) of your examples tell a story, beyond a one-word description or simple phrase. #Cheers
Na that's not his intention for this video, like you said he wants to showcase the parameter, NOT to tell a story..
@@mariomills What ur sayin would be relevant if he were only doing 1 or 2 examples. But he often does 12 or more each video. There is ZERO reason why he can incorporate more narrative prompts. Also, some of these parameters work terribly with narrative (complex) prompts. I'd like to see him incorporate that. It's not an illogical ask.
serious storytellers shouldnt be using ai image gen right now anyway, its still in its infancy. use an established art form like 3d/digital painting/etc. ai still cant even handle multiple characters or objects interacting, which most stories have. good luck making the background have any coherency or the lighting to make sense too lmao. ai is great at improving art though, all the upscalers and controlnet are fabulous (midjourney has neither)
@@spacekitt.n I disagree. The most recent -sref and -cref prompts are a game changer. Stories can be told and told well (I’m working on a film pitch deck right now - something that, until, like, yesterday, you could only dowith stock footage unless you shot it yourself) It’s not perfect, but the current AI art technology in combination with the AI magic that is available in editing programs like Adobe Photoshop and/or Canva, you can produce work on a solo budget for chump change that would’ve previously cost tens of thousands of dollars. (also, it would be foolish for “serious storytellers” to not start learning Midjourney at an advanced level NOW, so that, as new parameters, such as the ones mentioned above become available, said storytellers, like me, can hit the ground running).
I really appreciate the feedback! The reason I use simple prompts is pretty straightforward - I want to showcase a feature using the least amount of variables as possible. Just as you're turned off by the simple examples, I worry that new users would be turned off by more complex examples. But I like your suggestion and I'll try to incorporate longer prompts into my videos!
alpha is dead since 2024.4.22 404
I had the same problem but then it randomly worked for me again lol so I'm not really sure what happened
me too