Very impressive tutorial. Loved your time travel video at the end. Even though there are obvious limits, the fact it is at the stage an entertaining storyline can be manifested is great. It involves a lot of work, though, at this stage, that is why won't see tons of people doing it.
@@aivideoschool yes, lots do not understand it is nearly like doing a cinematic cartoon or stop-animation; each pose has to be drilled at in order to get, and that can take several prompt tries, then from sometimes hundreds of short clips harvested - as it is almost like reaping crops, lol - then selecting what works, that also hit and miss sometimes, getting some unexpected gems, and well, hours and hours go by for just video minutes long. So, in a way, right now, that is the assurance there won't be a flood of decent videos as the one you did in your time travel story. I say this having turned one of my music videos into ai visuals, though I sang, wrote lyrics, music, and the instrumentation played by real human beings not purchased ai or loops. For that one you can see as my channel's featured video "Beware Fallen Angels" I generated hundreds of prompts, not knowing then about midjourney etc images to video, and even then I had to break down sometimes and make my own derivatives to 'fill the gap' of what the ai could not or would not do (over word misunderstandings). So I do appreciate perhaps even more what you did in your time travel, as well as all of your detailed ai instructions, as even though ai is lots of work, it can let persons create visions otherwise out of reach, so can be worth work for some projects. I just stumbled on your videos, but you have the best tutorials, and I thank you for making them!
New stack: - MJ - CapCut or Canva to cut out character - Animate background separately with Pixaverse (better more stable motion brush) or Stable Video - or just camera move with... - Lip Sync with Pika or Sync Labs - Optional cleanup step with Topaz or some of these ComfyUI steps
Thank you, great video! However, Pika doesn`t offer lip sync at the moment and runwayml didn`t work with my animation videos and it also didn`t work with faces in profile
Pika does offer Lip Sync, I just tried it last night. You have to upload an image or video, then you'll see the Lip Sync button. It might not work with animated characters. They have to look human-like for it to recognize the face.
Can you please let me know how to lip sync when there are 2 characters in the same scene? is it possible? Runway keeps detecting the wrong face. (I also have a robot in the scene which isnt getting detected at all)
The only way I know of currently is to cover/paint over one face in the photo and lip synch the first character. Then paint over the second face and lip synch the other character. In your editing program, crop or mask the two clips so you only see the talking faces.
P.S. You need to add 0:00 at the top of your chapters list in the description so that the chapter markers will show up on the videos timeline, otherwise they are just buried in your description area, hard to find and not helpful to viewers
@@aivideoschool I figured that. My pleasure my friend. By the way, in my hunt to find the best quality AI videos, I stumbled across your channel and I saw you reviewing a lot of the platforms that are already somewhat known. But I just stumbled across replikant. It's more of a 3D animation software, but I think it may be relevant to some of your followers as it allows ultra realistic storytelling videos to be created with their platform very easily without having to have years of 3D animation experience. It's a new platform made by epic games I believe.
Glad I showed up in your hunt. Replikant looks so cool but I don't have an Nvidia GPU (I'm on a Mac M2). I wish they had a cloud version like Wonder Dynamics but it looks a lot better than Wonder.
The best way I found prior to Midjourney's update was to uniquely name the character in the prompt and use a detailed physical description that includes age. Then you copy the seed number from an image you like. If you add onto the original prompt with the same seed, it's sort of similar. I don't think Dalle will give you a seed but Leonardo AI will and is also free.
@@aivideoschool thanks my friend. there's a new way to make ai movies you have to make a huge video about it which is creating a full body shot image for a character and his background and make the character move with viggale ai. this is much much better than the slight motion we have with runway, pika or haiper.
This is quite easily the best tutorial for consistent characters. Well done.
Thank you!
Very impressive tutorial. Loved your time travel video at the end. Even though there are obvious limits, the fact it is at the stage an entertaining storyline can be manifested is great. It involves a lot of work, though, at this stage, that is why won't see tons of people doing it.
Thanks! I'm glad you point out there's a lot of work involved. This gets underestimated by a lot of people.
@@aivideoschool yes, lots do not understand it is nearly like doing a cinematic cartoon or stop-animation; each pose has to be drilled at in order to get, and that can take several prompt tries, then from sometimes hundreds of short clips harvested - as it is almost like reaping crops, lol - then selecting what works, that also hit and miss sometimes, getting some unexpected gems, and well, hours and hours go by for just video minutes long. So, in a way, right now, that is the assurance there won't be a flood of decent videos as the one you did in your time travel story. I say this having turned one of my music videos into ai visuals, though I sang, wrote lyrics, music, and the instrumentation played by real human beings not purchased ai or loops. For that one you can see as my channel's featured video "Beware Fallen Angels" I generated hundreds of prompts, not knowing then about midjourney etc images to video, and even then I had to break down sometimes and make my own derivatives to 'fill the gap' of what the ai could not or would not do (over word misunderstandings). So I do appreciate perhaps even more what you did in your time travel, as well as all of your detailed ai instructions, as even though ai is lots of work, it can let persons create visions otherwise out of reach, so can be worth work for some projects. I just stumbled on your videos, but you have the best tutorials, and I thank you for making them!
New stack:
- MJ
- CapCut or Canva to cut out character
- Animate background separately with Pixaverse (better more stable motion brush) or Stable Video
- or just camera move with...
- Lip Sync with Pika or Sync Labs
- Optional cleanup step with Topaz or some of these ComfyUI steps
Very nice tutorial and an amazing end. Really appreciate the video.
Thank you for watching the whole thing, I hope it was helpful!
Great video!
Thank you, I appreciate it!
thanks Mike. this is great.
thanks for the tutorial. this is great
Glad it was helpful!
great video - thank you:)
awsome. Clear, concise, and easy to follow. Excellent visuals and examples. Very helpful. Thanks!
Excellent Videos 😎💯
Thanks for watching!
17:09 LOL this is AWESOME!! Though I prefer Runway's lipsync because it has facial movement as well.
Good tutorial
Thank you, great video! However, Pika doesn`t offer lip sync at the moment and runwayml didn`t work with my animation videos and it also didn`t work with faces in profile
Pika does offer Lip Sync, I just tried it last night. You have to upload an image or video, then you'll see the Lip Sync button. It might not work with animated characters. They have to look human-like for it to recognize the face.
Can you please let me know how to lip sync when there are 2 characters in the same scene? is it possible? Runway keeps detecting the wrong face. (I also have a robot in the scene which isnt getting detected at all)
The only way I know of currently is to cover/paint over one face in the photo and lip synch the first character. Then paint over the second face and lip synch the other character. In your editing program, crop or mask the two clips so you only see the talking faces.
P.S. You need to add 0:00 at the top of your chapters list in the description so that the chapter markers will show up on the videos timeline, otherwise they are just buried in your description area, hard to find and not helpful to viewers
Thank you for this tip! I didn't realize that was the trick to get them to show up in the timeline.
@@aivideoschool I figured that. My pleasure my friend. By the way, in my hunt to find the best quality AI videos, I stumbled across your channel and I saw you reviewing a lot of the platforms that are already somewhat known. But I just stumbled across replikant. It's more of a 3D animation software, but I think it may be relevant to some of your followers as it allows ultra realistic storytelling videos to be created with their platform very easily without having to have years of 3D animation experience. It's a new platform made by epic games I believe.
Glad I showed up in your hunt. Replikant looks so cool but I don't have an Nvidia GPU (I'm on a Mac M2). I wish they had a cloud version like Wonder Dynamics but it looks a lot better than Wonder.
for those who can't afford midjourney. Is there a way for consistent characters with free Dalle-3 inside Microsoft bing. thank you very much.
The best way I found prior to Midjourney's update was to uniquely name the character in the prompt and use a detailed physical description that includes age. Then you copy the seed number from an image you like. If you add onto the original prompt with the same seed, it's sort of similar. I don't think Dalle will give you a seed but Leonardo AI will and is also free.
@@aivideoschool thanks my friend. there's a new way to make ai movies you have to make a huge video about it which is creating a full body shot image for a character and his background and make the character move with viggale ai. this is much much better than the slight motion we have with runway, pika or haiper.