Thank you Tao for keeping us informed about the latest developments with Kling AI and other tools like Minimax. I have learned a great deal from your tutorials and appreciate the clarity with which you demonstrate how to use these tools. Your hard work is greatly appreciated. Please continue creating content, as capturing our audience’s attention with beautiful and high-quality content is what truly matters, and you deserve all the recognition for it.
THIS - IS - INCREDIBLE!! Thank you so much for letting us know and the clear instructions...the two programs I'm using now do not allow for prompting...and the bodies remain stiff about 85% of the time, so this is exciting!! I will try it out this weekend. Thank you so very much Tao!!
Thank you! Yeah, Kling is doing some really cool stuff right now. You can make pretty good quality short videos by combing Kling, Runway and all the other new platforms coming out
With the 2 person video: to get the woman to speak perhaps you could edit the source video, blur out the mans face (with tracking) and then use that to force the AI to detect and use the womans face (which should be dominant) - then use the output to combo-edit the clip with the same of the man speaking. Just an idea - I havent tried it :)
Thank you Tao, it was a wonderful video. You answered all the questions I had in my mind, I have only one question. What should we do if we want to show someone's feelings? For example, someone is angry and shouting or she is sad and crying?
Glad this helped with your questions 👍. To show facial expressions, I would generate those inside the "base video" before applying the lip sync. Once you have that, the lip sync can be added on top
Very neat, but still prefer LipDub AI as it allows you to choose the actor or actors that you want to lip sync in the scene, and provides Hollywood quality lip-sync
Different approach... you can use kling AI to make your training footage to use in LipDub. You need a 1min footage to train LipDub. It doesn't make the first lipsync if you were using an image without mouth movement. One does not substitute the other.
@@taoprompts I just did my interview with LipDub's Andrew. I got access to Beta. I will now make a really crazy video soon! I want it to be crazy good so itll prolly take me a week to produce.
Thanks for all your great videos. I tried it immediately. It converted naturally even if the microphone was somewhat covered by my mouth. Too bad I can't extend it. I guess the only way to do it now is to generate it again at the last frame and connect it?
Yeah I think that would be the best way to do it. Kling 1.5 doesn't support extending videos yet, although version 1.0 does let you extend 5 extra seconds
I think Kling is better for a single character. But if you have multiple characters in the same shot, runway does give you more control over which character you want to lip sync with.
Hello, I just saw pika update 1.5 on video, it looks crazy. We can put quite realistic effects. Object explosion, inflation effect etc... Ps: I made shorts with kling ai, it's my last short, don't hesitate to give me strength, thank you 🙏🙏🙏
with all these video generators and kling being around, do we even need sora? i mean when they first showed it it was something never seen before but instead of just releasing it they chose to keep working on it so it better be something groundbreaking
Can you use any video you want or even an image or does it have to be a clip generated with their tech? Also what is the maximum length of syncing? Thanks.
Do you know a way to stop the aperture from changing when the light changes I find that if there are light changes, it's like the camera is stopping up and stopping down. I've tried negative prompts but no luck.
@@taoprompts Thanks, I have found some other similar information from "NVIDIA's Generative AI NIM Agent Blueprints for Developers" on RU-vid. You should also check it, it's interesting
Thanks. Kling is looking like a good option, but there is no payment option with PayPal or Apple, etc. No way, no how, am I submitting my credit card info to an overseas company in China. I hope they change that.
@@taoprompts Thanks. I just watched a good compassion between Runway, Kling, PixVerse and Luma. Kling kept coming out on top with Runway kissing it. However, unlike Kling, Runway DOES have the unlimited monthly which means, getting more prompts than Kling. I had some decent results with Runway's trial and will consider it. ✌
hello, can you tell me which paid version of kling AI is perfect to buy as a beginner.....but let me clear you that i am planning to create 10 minute videos like short films with dialogue deliveries in hindi.
For a video that long, I would purchase the $80/month plan if you use professional generation model which I do. If you use standard generation mode, a cheaper plan will work
Heres the problem with 100% of all these lip sync attempting AI platforms. Voiceover work as we are all used to from hollywood productions is not just lip sync, its ambiance/environment tone and the characters voice, energy, delivery and cadence must match as well. This is why it always sounds OFF. Theres also a misalignment of Mic proximity that throws everything off. These AI voices all sound Podcasty or Audiobookish. They do not sound like a boom mic thats a few feet away from the actual actor whos onset. Voice overdubbing is an Art that cannot be easily replicated by AI, it will take some time for us to get there. but in due time maybe
Yes the Ai voices do sound a bit out of place, but the tech is pretty amazing. I think it will be easier to correct the sound than to animate the lips and facial movements
@@taoprompts that is true. As far as adding background, environment, noise, room Echo, etc, but even after that, the cadence and other little nuances will never match that easily. We have to remember that as humans, we are absolute experts at seeing facial expressions and lip sync all day every day of our lives. So even the slightest deviation and it seems uncanny valley. I do have to admit, however, seeing what AI has been able to accomplish over these past couple of years, I think we will definitely get there soon where facial expressions, cadence, room tone, ambience and everything else needed to make lip sync believable from an AI character will be in our fingertips before we know it.
Hey Tao. How are you today my friend. I've bought a premium subscription for for kling but I found out that it's videos are like slow motion unlike minimax videos which appears natural motion like normal movies. I want to make realistic normal videos not slow motion. Is there a fix for this?
Hey, did you try using Kling version 1.5? I find that the motion is faster than in version 1.0. Sometimes I will speed the videos up a bit in post processing maybe 1.2 - 1.5x speed.
