NEW Midjourney Update with (old) Features
Вставка
- Опубліковано 31 тра 2024
- Midjourney Version 6 Update! We now have access to Zooming out, Panning, In-Painting, AND a new community feedback feature. Let me know what you think!
More Prompts ➡️ ftp.live/patreon
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
0:27 New Pan Feature
2:11 Zoom Out
2:59 In-Painting
4:01 Solving Common Problems
5:29 New Algorithm?
6:10 Feedback System
Please ignore the cut on my face. I'm not going to put on makeup - the show must go on
Should have just used the new in-painting feature on it 😂
Don’t even worry bro you good !
We could al do with a touch up before going on camera.. Some of us Need a mask!!
ahahahhaha @@conhuir it's a good one ahah
First rule of Fightclub: Nobody talks about Fightclub.
The feedback option was a nice touch from Midjourney, hopefully they listen to us and give us full CHARACTER CONSISTENCY :)
hahah go let them know how important it is
I did ahaha thanks to your great reminder with the ''feedback'' option :) @@FutureTechPilot
We also need digimon fusion references features ect sliders/prompts
Fusing other pics into wat we want to the system to turn those bundled combo genre pics into a newer version of our own ideal thoughts ect 🙏🏽👌🤪👍
Amazing! All I need now is Tune to be added to v6 & I'll be a happy little chap 😊
and then onto version 7 haha
LOVE these new (old) toys for V6. now... I just wish that Magnific gave any sort of beta credits. its fun to work on designs and then pan, upscale, pan, upscale, pan, upscale, pan, upscale but since we now lose all of that stitchable resolution over previous methods the only legit way I know to blow up these finished results with compelling content would be Magnific. RIP midjourney resolution, long live the king. ;)
Yah I hope we see some sort of advanced upscaler from Midjourney eventually!
This is cool. MJ is makeing good progress. Very exciting.
Oh yah, this technology is so fun
Thanks for the video! 👍Panning is so much better. I didn't see the rate other people's feedback. Will try again I suppose.
I didn't see the part to rate others feedback either but I just didn't realize you could scroll down on that page
Thanks Nolan, loved it! Finally these features are out, great, thanks for letting us know! I love the improvement of the pan, I didn't find it useful before because of exactly the effect that you pointed out. Now it seems to be what it should be! Can't wait to fool around with it 😉
Yes and please let me know if you find any cool ways to use it!
so glad these options are back!!
Yah it's definitely the best way to experience Midjourney
@@FutureTechPilot can i send you a prompt and can you explain why the hair is doing what its doing? it puts purple hair over a blond hair it seems??
Wow! That video has solved some of my queries i had. Your video tutorial is next level! You are doing excellent job! Thank you for sharing these tips with us. I was about to ask you how can we fix deformed hands/arms, animal legs, faces, from our old generated images, but I guess you have answered that in your video with the Vary Region option. Also thats awesome about the New feedback feature. I wanted to write a feedback to Midjourney but I didn't know where to write it so the developers will read it and take it into consideration. I want Midjourney to have a better understanding and include all of our words in the prompt. until now was ignoring many of our words. Where Dalle can understand our prompts better. and second, I want to be able to generate countries flags, last year Adobe firefly could generate my country Cyprus flag 🇨🇾 but Midjourney was only generating for me a piece of cloth, and a symbol that was similar to Lebanon flag. 🤦 Also During last Christmas I tried to generate an angel in the manger, with the (vary region) , and it was generated for me just a plain wing. 🤦 Midjourney needs a lot of improvement. It has weekness when we create a "group" of objects, or people, or animals, it messes up. I have a request. Do you know any other Ai tools that we can fix these image mistakes? for weird hands/arms, faces, animal legs and heads?
Well I really appreciate the kind words, thanks a lot! And a tool to fix image mistakes? I don't know of any a.i that specializes in that. I think your best bet would be to use Photoshop to manually fix any weird faces/limbs. Or even try Photoshops gen-fill a.i - it's really powerful in certain situations
The pan up/down feature works great. I made a full body shot out of a head shot in a few steps... really came out great without changing the prompt. When I tried panning left and right... I did have to change the prompt otherwise it would produce strange results, However... with a bit of playing around... I was able to create a really interesting scene from that head shot. This feature is greatly improved since v5.2 Thanks for the tip about using Vary Region on the black bars. I have some old images with these that I was not able to pan with before. I will revisit them again with these new tools and techniques.
Yah Pan can act a little funny sometimes but I'm really enjoying it so far!
Thank you for the tips:)
Cheers!
Love your channel Nolan....thanks
I'm really glad you enjoy the videos I make! Thanks for taking the time to comment
Thank you Nolan! Another great update vid. Why you don't have a million subscribers, I'll never know. You deserve it with your excellent content.
😂 just trying my best
Bro, SUPER useful and concise video, thank you
Thanks for the feedback!
Yes!! I SO missed this, now can make some infinite zoom vids w MJ6 🎉🎉🎉
Yah it's so nice having these options back haha
Great video!
Thanks for the feedback!
Thanks! I was so glad that this morning everything was much faster than the last few days. I’m trying to moderate myself, I’m happily addicted.
hahah I know what you mean
AWESOME UPDATE. using vary region to fix text that is just slightly wrong is a bing killer.
Oh yah!! Text. Jeeze. how did I forget about that. Thanks for the reminder!
thank you ☮
happy to help!
Thanks for the video, and exactly the same priority over the feedback as you
I'm glad we're on the same page!
Thank you so much for your hard work, your videos are very helpful.
And thanks a lot for taking the time to leave a comment. I really appreciate it!
They need to bring an option to be able to zoom in also.. cuz sometimes you zoom out too much and you’ve changed parts of the picture and just want to zoom in..
Yah some sort of slider would be pretty cool!
I think you just penned Midjourney's motto. "I just got goosebumps". 😆
😂 facts
(0:10) Technically, that's "outpainting", not "inpainting". The latter would be if you could highlight the roof of the car and convert it into a convertible.
You can do that with 'Vary Region' for sure ... probably lol
Great! A welcome update... excited for the future.
yes!
Thanks for the video. Do you know when all those features are available on a regular website?
I'll guess like 3 more weeks!
Video idea: With Vary Region, See if you can go in to older images and "fix" the text.
haha okay I'll take a look!
Hello, I love your midjourney channel. I wanted to ask if you know if it's possible to take a photo of an article with a white background and have it placed in a room or on a table with a prompt, without modifying the article itself. Do you know if this can be done or if there's any software that can add shadows to the object and position it where I specify with a prompt? Thanks, umma.
Unfortunately Midjourney can't do that right now. Maybe in the future. But for now, you could generate the backgrounds in Midjourney and then use something like Photoshop to place the object in the scene
I knew there would be 'extras' with these updates also knew you would have them sussed. As always thanks for your efforts man, great job.
Cheers!
I just knew there were 'extras' to be had with these updates, also that you would have them sussed. Great work man!
☮
🤝
Gosh I always forget to use the ''stylize'' and ''chaos'' features. I always use the size / zoom / seed / image weight but I always forget to the use the other two in my prompts. I NEED to start using them. Then again it's so many things to take into account ahahahah prompts making is like science ahahah. I always tell my friends that A.I doesn't ''do it all by itself'' we need to be the ''engineers'' of it when we use it. Thanks again for this great video :)
😂 it's impossible to remember everything in Midjourney
true ahaha there are sooo many things yet it's what it makes so much fun@@FutureTechPilot
Thanks Nolan! I really have troubles using vary region, when it comes to erase something that you don’t want. I tried by modifying the remix prompt, changing the whole prompt (as you did with the lightsaber), using the -no command, but still don’t work. Any advice ?
Hmm .. I don't think it's great for inserting/removing objects, but it's really good at replacing objects - like turning a sword into a lightsaber. If you need to remove something, Photoshop is 10x easier and legit 10x better at it. It's actual magic
I’ve been needing these updates.. I hate how it disappeared
they needed time to create them for V6
Super
🥇
Thank God
Hello I am seeing a change on midjourney display. Because when I click on one image I have generated in the midjourney website it suggests me other related pictures. And sometimes It helps to see the other people prompt. But now nothing. Also the explore page is also limited. When you search and find one good image and want to see similar there is no more suggestions. Is that only from my side? Or is it to protect other users generated images?
I noticed that too but I think it's probably just a bug
🔥🔥
🤝
There is a limit of 140 characters on the /feedback field. Do you think that is enough if you want to suggest with specific details?
Yah I guess it needs to be easily digestible - be as specific as you can
Probably important to note that these features are only available on the Discord Server. Pan seems to be available on the alpha website (I didn't test it), but inpainting is not available on the website. We may have to use different "species" of Midjourney depending on what we want to accomplish.
oh dang! Frustrating for sure, but I know it's a high priority to get the website feature complete
do you have a v6 course?
Working on it!
Does anyone know of a work around for how to remaster images from v5.2 into v6? Thanks
You can turn remix on and hit a variation button, then change the version to --v6 ... or you can include your 5.2 image as an image prompt, set an image weight (--iw 2 or --iw 3) and prompt in v6
hope that helps a little
@@FutureTechPilot superrrr, thanks 🙌
Wait you can finally zoom out to a specific AR? I make a lot of phone backgrounds and have been asking for this feature for months😁. The problem with generating in 9:19 AR is all the outputs are very very similar and always super symmetrical. So now I can generate in 1:1 or 16:9 AR for ideally more variance and then zoom to 9:19. Wait can you zoom in???
😂 no, no zooming in. not in the way you're thinking - but you can kinda fake it by starting really zoomed in, and then zooming out haha
@@FutureTechPilot lol yah I was a little disappointed when I found out. Definitely works a little better though cheesing it
im trying to upscale to 4k since the update from v6 i havent been able to figure it out. any ideas?
They haven't added the 2x and 4x upscalers to V6 yet - I think we'll see them eventually. For now you might want to try a software like Topaz Photo AI - I'm not shilling for them, I just think their upscaler is amazing. You could also try Upscayl - www.upscayl.org/ - you can download it for free and it runs off your computer
I am happy for the zoom. I just had to switch to 5.2 so I could use the feature.
Yah it's a lot of fun!
Tell me, after the updates I stopped getting realistic pictures of people’s faces. became more plastic. Maybe something needs to be added in the prompt now?
It's a good question. I'm not entirely sure, please let me know if you find any good words to use. But I would suggest trying --style raw and maybe lowering the stylize value --s 30 - 80
The vary region doesn't work well for me. They still need to do some work...
Yah Photoshop Gen Fill is way better
Nolan, any idea why they decided to change the upscale tools? Version 5 upscale was perfect.. now it constantly changes the image, makes it smooth in many cases.. really annoying. They had nailed it in the previous version.. why change something that works perfect.. 😣
They have to remake each tool when they update the version, and they still might add those 2x and 4x upscalers eventually. You can check out something like Upscayl though - www.upscayl.org/ - it's free to download and runs off your computer
Can you upload your own picture and in paint from there?
no, but maybe in the future!
/imagine Future Tech Pilot -- no cut on face --ar 3:5 --v 6
😂 I wish
In my opinion, intelligent beings comprehend the subject of artificial intelligence, but for many, it seems incomprehensible. You've created many excellent videos; I'm eager to see more of these examples and become acquainted with the future sooner. By the way, this text will be translated into English by an artificial intelligence.
haha well thank you for going through the effort to leave a comment!
think the volume was way too low on this one lol
Okay thanks for the heads up!
inpainting kinda weak for V6 😢 but i have hope
Yah let's keep our fingers crossed!
Omg i almost wet my pants with this update
😂
My image no longer seamless when I used the pan
hmm, maybe you should ask around in the 'support' channels on Discord! Sounds frustrating for sure
Great. Thanks. Just a little correction: It is impossible to pan up or down, I'm afraid. You can only pan left or right, as the word "pan" comes from "panorama". But you can TILT up and down :) I understand this misunderstanding is spreading, though.
😂 #technicallycorrect