One cool thing an amateur animator could do is having midjourney generate line art in the form of pencil sketches for your characters, then animate them as you described here. Then just clean up the line art as you would in the traditional animator workflow. AI did most of the keyframing and in-betweening, and now the human comes in to tidy up the work and shove the artifacts under the rug!
I've been wanting to make some fun videos of characters dancing, sort of like the dancing cats video I just made but with static backgrounds. I wasn't sure how to keep the backgrounds consistent and didn't want them morphing all over the place. This should be a good way to do it, although I might use Photoshop to cut the characters out from the background manually. Thanks for another great video!
Good to hear this was helpful 👍. Luma does still struggle with animating complex motions without blurring, but I think for dancing cats it should work pretty well.
@@Warmer_Bros I just re-read your comment. I also think it’s weird that it added the random lip movements. Maybe negative prompting might help to remove it.
I have noticed that @1:25 the motions cant be too huge. Ive used it to manually control facial motions or simple stuff like a character fallling to the ground or pointing their finger at their head etc
One cool thing an amateur animator could do is having midjourney generate line art in the form of pencil sketches for your characters, then animate them as you described here. Then just clean up the line art as you would in the traditional animator workflow. AI did most of the keyframing and in-betweening, and now the human comes in to tidy up the work and shove the artifacts under the rug!
That would be pretty cool, there's a lot of fun animations you could make with that.
Awesome. Every time I start to get bored with AI I just have to come to this channel to get inspired.
Thanks Val, there should be a lot of big Ai updates in the rest of 2024 👍
Tao! Question for you! Does Luma have negative prompting?
Thanks for another great tutorial!
Geez Tao...THANK YOU!!
I've been wanting to make some fun videos of characters dancing, sort of like the dancing cats video I just made but with static backgrounds. I wasn't sure how to keep the backgrounds consistent and didn't want them morphing all over the place. This should be a good way to do it, although I might use Photoshop to cut the characters out from the background manually. Thanks for another great video!
Good to hear this was helpful 👍. Luma does still struggle with animating complex motions without blurring, but I think for dancing cats it should work pretty well.
Now this is cool
Thanks man 🙏
your videos are amazing!
Thanks, I really appreciate that!
Thank you, great👏
Good idea dude.😊
So awesome!
Thanks man!
Its amazing, is there anyway not to animate lip movement? Can you make a tutorial.
Haven't tried, but putting Lip movement as a negative prompt might help
@@Warmer_Bros We want lip-syncing to match up with the words that the character is saying, not just random lip movement.
@@Warmer_Bros I just re-read your comment. I also think it’s weird that it added the random lip movements. Maybe negative prompting might help to remove it.
@@Warmer_Bros thanks a lot 😊
You could try Hedra AI or lip sync from Sync labs.
Thanks!!!
Awesome ❤
Please can you do a tutorial on how to make book cover using midjourney 😢
I will start making Midjourney guides again, I'm excited for the 6.5 release
Thx ❤
I have noticed that @1:25 the motions cant be too huge. Ive used it to manually control facial motions or simple stuff like a character fallling to the ground or pointing their finger at their head etc
Yeah, this works best for smaller motions and expressions. I haven't seen any Ai video be able to generate complex Anime motions yet.
Just noticed there's now a loop option on Luma Dream Machine. Did you see that yet?
Yeah, I posted a video about it. It works really well with a bunch of different visual styles
ToonCrafter but way better. This is the way if you dont want to fumble around with ComfyUI just yet (or are procrastinating learning like myself 😅)
is there an uncensored version of this?
You'd most likely have to go for open source models
Not with Luma, you would need to set up something like stable video diffusion locally to get uncensored videos.
still's jank af, waiting for kling to drop - they seem to have a way better algo
Free ai tools?
Kling AI has a free version, although it is slow
Eh good luck