2 ай бұрын
3 ай бұрын
4 ай бұрын
6 ай бұрын
6 ай бұрын
7 ай бұрын
9 ай бұрын
9 ай бұрын
10 ай бұрын
11 ай бұрын
Пікірлер
@Mouriyaraja 12 сағат бұрын
Hi huge fan from India, I just want to request one doubt alone. Which AI video generating software gives best character consistency, please let me know sir... It will be a really great help for me...
@Gales_of_Gaelic Күн бұрын
I went to Pinokio and it wants me to pay. Thought it was free. suggest?
@JM-wt3dc Күн бұрын
i don't get the manager button after instaling it :(. What am i doing wrong
@1BrknHrtdRomeo Күн бұрын
I honestly got better results off of Pixverse?
@ValentinesCreatives 3 күн бұрын
Thanks! Hey, do you have any videos on using civitai to make AI-generated images of oneself without having to install any programs?
@metapolys 4 күн бұрын
It helped me a lot, thank you very much for sharing.
@BlackMixture 3 күн бұрын
You're welcome! Glad it helped!
@Haligonian 4 күн бұрын
Awesome stuff!
@HarrisonsFord 6 күн бұрын
One suggestion, putting links to all the files you mention in the video in the description would be useful! It is easy enough to search for the comfyui github, but a list of links would be nice.
@KingBop313 6 күн бұрын
so I trained and it said training complete but nothing pops up. I do see stuff in my files however if I go to my c drive then pinokio then outputs --> toml files. However how do I open them in forge? Video doesn't show that step
@大支爺 6 күн бұрын
The target object CAN NOT has any movement.
@103zbq 6 күн бұрын
This workflow keeps on bringing more and more troubles rather than working. Would you be able to get it to work?
@John-mn3iu 6 күн бұрын
That first clip reminded me of george droid fent memes 😂
@boyboy168 7 күн бұрын
I tried this function with my charactor, but it created some creepy clips, which I am not sure they are usable. And I think the core in this function is text-to video , which I seldom use, as it could not give consistent background and style.
@TopFactology 7 күн бұрын
The thing I dont like with minimax is the faces, surfaces seem airbrushed, no texture…
@cinematictrailerstudios 7 күн бұрын
We need to go back to gate keeping giving everyone the sauce makes it not unique no more lol then we get 1000 low quality videos on youtube
@Siraj-d1n 7 күн бұрын
Thanks Nate will share your vid
@JackRealy2001 7 күн бұрын
no one has the courage to say the real things; we don't want to lose our jobs and we want to forcibly boycott the ai! let's not waste time saying that humans will always be the best and other comforting phrases.. we simply have to stop the AI! the same goes for newsstands, pizza sellers, waiters.. it may also be evolution but we must stop it.. we must not make philosophical speeches.. we are afraid and we must defend ourselves
@JackRealy2001 7 күн бұрын
nobody is scared? no actors,no musician....how can i help me my mind?
@jazzampa 7 күн бұрын
Musicians are indispensable
@ModRebelMockups 7 күн бұрын
it asked me for a trigger word. do i just pick whatever word? im confused when i would use that word because in your vids i see you only use the two names u add to the captions. u never show what you put as your trigger word or how to use it
@apatsa_basiteni 8 күн бұрын
I give the tool 5/10
@thewebstylist 8 күн бұрын
Very cool, I just don’t think they should charge the credits whenever a wonky unusable vid is produced (my issue w other vid platforms) Great vid
@georgehephzibah4786 8 күн бұрын
I LOVE YOU SIR....
@HarryClipzFilmz 8 күн бұрын
You look like Big Boi from OutKast. I will give this another try since this before this
@TheManInCommand 6 күн бұрын
I always thought the same thing but didn't want to be rude lol
@HarryClipzFilmz 4 күн бұрын
@@TheManInCommand It's true in a good way
@pdjay8912 8 күн бұрын
구독했어요! :)
@LeniviyArt 8 күн бұрын
Thank you for sharing! <3
@rajsneh3 9 күн бұрын
Thanks for a great video. I love it. Just 1 question though, can you create image to video using this tool? If so, can you post a video on that or can you message me on that?
@_its.just.regi_ 9 күн бұрын
I don’t think it is the length of your prompt, but the colloquialisms used. “Puts on” for example may be confusing the model even when it is trained to understand natural language tokens. I would describe any actions in a more precise and formal way leaning more towards technical or descriptive contexts.
@viktorsz 9 күн бұрын
cannot wait a platfor where you can modify and change details attributes..so far Kling is the winner for me
@VfxKopele 6 күн бұрын
You will have to work with the developer versions of these platforms to fine tune the generations.
@OBthaBull 9 күн бұрын
With my phone, can you provide instructions on how to take the highest quality image or can you let me know which video you’ve already trained us on this? Thanks.
@daily-quranic 9 күн бұрын
First!
@Galamboimagesearch 9 күн бұрын
Turning photos into videos has never been this easy! 🎥✨ KLING AI is truly impressive. Looking for more AI-driven magic? Explore Galambo for seamless image search solutions! 🚀 #Galambo
@ModRebelMockups 10 күн бұрын
could i follow this video to do a photo with more than 2 people you think, maybe 4-5 people? or a diff approach would be needed?
@BlackMixture 9 күн бұрын
So far I've only trained a lora with two people max, but I have been able to do 3-4 person scenes with two loras combined in one workflow. There might be a better method though because sometimes the style of one person may dominant the image and a lot of fine tuning the weights means you'll have to generate a bunch of images before you get the perfect settings.
@ModRebelMockups 10 күн бұрын
would you be willing to let us view your dataset of images? I want to make sure my dataset is varied enough and similar to yours so I can get results like yours. Im sure the dataset matters a ton. Even just low res images, for pose reference.
@ModRebelMockups 10 күн бұрын
if you had to guess, what is minimum number photos per person, and num photos with both of you in them, to put in the training data set?
@BlackMixture 9 күн бұрын
I think a good data set can be as little as 17 images. You can have both characters together in half of the data set (8-9 pictures) and then individual photos of the characters in (3-4) images each. Flux is very smart and can understand a lot with very little guidance. Our images were mostly selfie photos, photos taken when traveling, and a few super random solo photos. Most of the data set was shot on our android phones so they're not the highest quality photos.
@artekal3d 10 күн бұрын
wasn't expecting MKBHD one to turn out like that, it was better than I thought! I noticed that the funko pop one at 2:40 had 360 output a bit off-axis. I thought maybe it was because of it off-centered in the image but I checked there's little to no off-centering in the image. Maybe the fact that the legs positions not being symmetrical did that? But that's still great overall!
@nondor6852 10 күн бұрын
great video no question about it, but why bother uploading it here, saying that comfyUI is great bc its free and then making the main example on how to use it on a workflow that is locked behind your patreon paywall? I tried following along with the free v2 version but unfortunately that didnt help much since its a lot different than the v3 version
@BlackMixture 10 күн бұрын
This is a follow up on previous video actually. We released the V2 version for free and a lot of people ended up asking how to install ComfyUI. This video should help you understand the V2 version: kzbin.info/www/bejne/g36QpHikaL5mlbM
@sdgtr4 11 күн бұрын
free as 4090ti
@BlackMixture 10 күн бұрын
Lol, you can also run this on runpod which let's you rent gpus mad cheap. I think a 4090 is about 35 cents an hour. We're working on a video about using runpod since a lot of people don't have crazy 4090s installed on their home machines.
@PIXALRIOTS 11 күн бұрын
No dice for Mac 15
@PIXALRIOTS 11 күн бұрын
does it not download cloud flared for Mac OS 15?
@Elden_String 11 күн бұрын
Newbie in this world Currently i have an old macbook air from 2015 with a I7 core, so i know that, running Confy UI to generate Images and Videos is not possible on it Does it work well (and very well), on the last m3 or m4 Macbook ? Thank a lot, i have a lots of projects i dream to do with AI :O
@rikudmode 11 күн бұрын
how did you put the fire on your hand 25:41. Help plz
@kutsanan 13 күн бұрын
Can i axport as stl?
@mo_mo_meaw 13 күн бұрын
now what do you think after veo2?
@JohnSmith-u8h 14 күн бұрын
Really straight forward and easy to follow thank you
@NEWLOGY 15 күн бұрын
I installed Expression Editor in Comfy UI, but it shows me this error: " Error while deserializing header: MetadataIncompleteBuffer" Please help me. It's very necessary.
@EpicNarrationTales-i3j 15 күн бұрын
hi i subscribed to your patreon, but im now having issues since i runed the advanced workflow, im not getting the preview loading in the sample custom advanced section now, just the green loading bit, what could go wrong? this now happend to both advanced and basic versions. as with the preview i get a rough idea how it will turn out, but i have to wait till it renders which wastes time. any help appreciated thanks.
@BlackMixture 15 күн бұрын
Hey thanks for subscribing! Send me a DM on Patreon so I can help you figure out what going on with your setup. My first guess is that If you've updated your ComfyUI then your preview settings might have changed and need to be adjusted.
@EpicNarrationTales-i3j 15 күн бұрын
@@BlackMixture reply sent now
@EpicNarrationTales-i3j 15 күн бұрын
thank you it fixed it. One more question with the version 3 workflow. Can i keep the background as it is without changing it? i just want to add my lora and change the character on the image with the same pose, but i notice i get the perfect pose copy with my lora but the background changes unless i prompt something similar(WHICH I DONT WANT) please let me know how i can use the mask feature if this helps me acheive what i want> also what can i do with enable mask options? do you have a written guide?
@nyvids9706 15 күн бұрын
Can you combine multiple Loras?
@buschlen 16 күн бұрын
Do you have a tutorial on how to do this using RunPod?
@BlackMixture 15 күн бұрын
Not yet, but working on it!
@ondrejzuna6070 17 күн бұрын
Hi i downloaded your workflow from patreon, but i cannot solve this error: Any help please? Thanks for amazing content! Failed to validate prompt for output 221: * BasicGuider 234: - Required input is missing: model
@BlackMixture 15 күн бұрын
Hi this error is related to using an incorrect model. Have you double checked you're using the Flux model linked on the Patreon? Feel free to send me a DM on Patreon and I can help you figure out what's going on with your setup.
@emmettfrancis 17 күн бұрын
Can this be done with 360 images or video? 😊
@TheGr00m 17 күн бұрын
I've been seeing vids about gaussian splatting lately, it seems to be the big new 3D thing. I don't understand though, how is it different from photogrammetry? Apart from the fact that it doesn't actually build real geometry?