5 New Things Soon | Midjourney Office Hours Recap Aug 14th 2024
ฝัง
- เผยแพร่เมื่อ 10 ก.ย. 2024
- NEW Midjourney Advanced Color Course ➡️ ftpai.thinkifi...
Midjourney Beginner Course ➡️ ftp.live/mjcourse
1-on-1 Consultation ➡️ calendly.com/f...
Prompt used: manga cover, title "___", [subject] in an action pose, in the style of Katsuhiro Otomo and Jean Giraud, bright colors, ultra detailed, best quality, high resolution --ar 47:64 --c 3
- Special thanks to stacybeatrice549110 !!
► FREE PDFS
Prompt Glossary V6 ➡️ ftp.live/V6Glo...
2-4-2 V6 ➡️ ftp.live/242
200 Aesthetics ➡️ ftp.live/200Styles
Midjourney Cheat Sheet ➡️ ftp.live/CheatS...
► SOCIAL
Discord - / discord
Instagram: / futuretechpilot
Twitter: x.com/FutureTe...
I wish Midjourney would add the ability to *‘combine multiple character references together,’* as well as *ControlNet to copy a pose from an image.*
I think that will happen one day!
Isn't that what depth control net will do? Or does depth control not copy poses?
I could care less about video and 3D and care more about making the quality and character consistency better.
I agree 100% with you. I feel overwhelmed. The documentation is not really updated for the website. Also accuracy, consistency and quality text should now be 100% working but it isn't. The heck with advanced features if they can't get the basic features to work correctly.
I think everything will keep improving over time!
16 images in a gen would be great… especially when you combine it with CHAOS… then i wouldn’t have to always do “- repeat 5” to do a bunch of rolls to get what i want and eat up my hours… will make it even faster and cheaper to get what you really want 💪
😂 exactly - I'm scared of how many pictures I'm going to end up downloading
I have an idea for an option within Repainting that I think would be extremely useful. Mirror symmetry is what I describe it as. Select the area you want changed, select mirror symmetry, have an exact same size area pop up (like using Clone in Lightroom) but flipped, drag it over the area that you want to mirror and then voila, MJ repaints it. I'd love to see light room include the same option with Clone where you could flip the selected area and mirror another area of the pic to get that perfect symmetry. What do you think?
Sounds a little complex for a Midjourney feature 😂 but you should let them know in one of the Feedback channels!
I'm happy with the amount they provide. Just viewing and sorting is a mess. I'm doing creative gen, loads of in-painting edits and also title treatment tests which are all in the same scroll with dozens of images that look the same if you're dealing with in-painting. Additionally, there's no way to search for in-painting images. I have like 40 images that all have the name "squirrel hands". I've posted more than once asking for something like "--Null Tagging" where we can enter in prompt text that's not processed by the engine, but searchable. Like I could ad to that prompt "squirrel hero 27", then I'd know which one it is without adding 'hero' to the prompt engine without it adding the number 27 or a cape or something. Thanks Nolan!
Yeah some kind of tagging system would be awesome - or just a 'place' to generate for specific projects - anything to make it easier when dealing with large amounts of images!
@@FutureTechPilot Seriously. Yeah some kind of sorting/sectioning through prompting would be invaluable. It gets discouraging when you can't organize. It feels more like work than fun. Fingers crossed.
Some great features on the burner and luv the prompt and art!
Oh heck yah
Helping out again. Succinct and all that. Way to go man............
Cheers pal!!
I was JUST debating suspending my MJ again, so I will hold off. I love new features.
haha let's hope it lives up to the hype!
I want to know what prompt u gave to generate the images that are there in his video I'm making a comic strip using midjourney please can you help?
It's in the description!! You can copy it from there
I just want consistent characters whether its a photo realistic image or artwork. Yes, I'm aware there are ways that this can be done but honestly, if I have to "work" at getting a character to be consistent in each generation in a ai image generator, I might as well try and teach myself to draw.
haha patience my friend! I think all these features will improve in the future
@FutureTechPilot lol yup. Tis why I've subscribed to your channel. Been keeping an eye out on the improvements. Thank you for your time in making these videos.
My question is how much longer do we have to wait for niji 6.1?
I don't think there will be a 6.1! They're working on Video
@@FutureTechPilot@FutureTechPilot When do you think niji will have an update of some type?
@@deanm9652 I'm really not sure!
Do you know if they're working on a Control Net feature?
Edit: Also I think it should be a togglable option (a switch) to choose if you get 4, 8 or 16 images because i feel it will be even harder trying to scroll up and down to find what you want. I know you can search keywords, but sometimes just scrolling takes less time.
haha yah I'm worried about my gallery filling up with images ... and they are working on something similar to control net - (not that I have any insider information) but I doubt it will work in the way people are expecting
Damn these images are amazing❤
Yah I was in love with the prompt when I first saw it!
I wonder what they have to say about Flux?
Competes with 6.1, not 6.2 = not a problem 😂
Competition is the best thing for all of us!
Deleting images from the website, that’s all I ask.
You can 'hide' images
@@FutureTechPilot not the same.
So very curious how 3D goes.
Yah I wonder what it will actually look like!
Sorry, but when you talked about "BATCHES" I was totally lost. Can you clarify or give a better example of what BATCHES are?
he’s talking about how you get 4 images for each gen
oops, sorry! Yah right now you get a batch of 4 images with every prompt - they're going to raise that to at least 8 but maybe even 16
Midjourney quality its good..problem with its same skin and faces..like its problem with how they train data..like flux its 12 billion parameter but they have diversity in faces and body
Try lowering the --stylize value!
Second 🤪
Really nice image this time, btw.
haha I'm glad you like them too!
I'm first:-)
🥇 You've won the Gold Medal today
@@FutureTechPilot YES!!!