Tip 2 - Animal Hybrid Fusion Prompt ( @officialmrlemon ): "zebra and a car leap into each other, colliding into each other. As they clash, they merge into a hybrid beast" Tip 4 Facial Expression Prompts: www.reddit.com/r/aivideo/comments/1g6shqk/minimax_expression_tutorial_imagetovideo/ Joy - Wide eyes, big smile, raised eyebrows, relaxed facial muscles. Sadness - Downturned mouth, furrowed brows, teary eyes, drooping eyelids. Surprise - Wide eyes, raised eyebrows, open mouth, slight head tilt. Anger - Furrowed brows, clenched jaw, flared nostrils, narrowed eyes. Fear - Wide eyes, slightly open mouth, tense brows, rigid jaw. Disgust - Scrunched nose, raised upper lip, furrowed brows, narrowed eyes. Confusion - One eyebrow raised, slightly open mouth, furrowed brows, head tilt. Contempt - One side of the mouth curled up, narrowed eyes, head slightly turned. Determination - Tightened lips, narrowed eyes, focused eyebrows, set jaw. Shock - Frozen expression with wide-open eyes, slackened jaw, raised eyebrows. Relief - Closed eyes, small smile, relaxed eyebrows, exhale shown with slightly open mouth. Embarrassment - Eyes looking downward, blushing cheeks, sheepish smile, raised shoulders. Tip 7 - Consistent Character Tutorial (FLUX): th-cam.com/video/5Z8fwEeWfRg/w-d-xo.html Tip 10 - Ink Illustration Prompt (@HalimAlrasihi): "Revealing an image from complete blank paper transitioning into the style of ultra-fluid seamless smooth ink brushstrokes that dually and effortlessly flow, merging like water on rice paper. Strokes appear soft and blend naturally, deepening in boldness with perfect consistency. The image forms organically, with each element revealing itself in a continuous, graceful transition." Full Tutorial: th-cam.com/video/XlT0vYGp1B8/w-d-xo.html
@@mariepaulecoutens 👍🏾👍🏾 Definitely, and it's worth it. There are lots of tricks and techniques to be discovered by experimenting and creating with A.I. tools. Best Wishes
Thanks, Bro! In MiniMax there is a trick I discovered for getting a consistent character by reuploading a still frame from a generated video with that character. How it works is that you upload that frame in image to video and prompt something like, "Fast blur transition to a new scene showing..." and then you describe that character in a new scene or angle, or even with new clothes. MiniMax then tries to recreate that character in a new scene. It may take several or more tries, but when you get another scene that looks perfect with the same character, you then download that video and use a single frame from it to then upload back into MiniMax to create a full 6-second clip with the character in that new scene. However, sometimes I have used a video that has had the prompt for "Fast blur transition" and simply spliced and cut off the first second or two when in my editing software, and then used the remaining 3 or 4 seconds (that show the character in a new scene) in my short film. This method CAN be a little time consuming, I admit, hehe, because sometimes it takes two dozen generations to get the perfect facial resemblance of the character in a new scene or angle or with completely different clothing (if that is prompted). I was going to make a video about this technique, but decided to share it with you here, but I may do a video on it next month. Cheers, Brother. Carlos
That's a super creative tip man! I've noticed Minimax sometimes pans to a new version of the same character, but never thought about using that as a way of getting consistent characters
Great teacher!! I think this video is like a gift to me~!! I haven't watched the video yet, but I will definitely apply for it after work today. Thank you :)😍😍
I absolutely LOVE Kling!! The price is reasonable, given all that it can do and the movements and details are phenomenal. I can't even imagine what Version 2.0 will bring.
Thank U! To explain well! I asked a question from a former video, so the solution is to describe it specifically. So is it difficult to make character in AI tools?
Tao - THANK YOU for all the great teaching. I am learning so much. I am struggling with one challenge and that is interacting hand movements. Using Kling with motion brush and prompts I just can't seem to get smooth natural movements. Also the hand placements always seem to miss where I show the motion going. For example placing a hand on an arm (bicep area) and it consistently goes to the chest or shoulder. In addition the skin as it is touched the AI makes it look leather like and not real. Any advice or perhaps a video tutorial would be helpful. BTW using Chatgpt has been a major help. Thanks again for all your teachings!
Hand motions are super hard to get correctly, especially between multiple people. I'd recommend trying a closeup shot on just the hand touching the bicep. So start with a reference image of a bicep and try animating a hand touching it
@@taoprompts So I tried your suggestion on Kling, Minimax, Pixverse & Luma. Kling was the only platform that handled it perfectly. But when I then included the full arm Kling, Minimax, Pixverse, Luma all missed it and went to the forearm ??? I even tried adding comments in the Negative Prompt input. And I even tried using Chatgpt to assist by showing it the image, prompt structure and generating prompts to try and each time the generation keeps goes to the forearm. Man I am so frustrated!! It would seem to be a rather simple task request why so difficult!?
I tried your suggestion and Kling AI worked perfectly, other platforms failed. However when I expanded the image to include the full arm Kling and others still placed the hand on the wrong location. I even tried Chatgpt using the prompt formats of each platform and still no success. I can't figure out why the models struggle with such a simple action? Any thoughts?
Hello there, I discovered your channel and I think your content is truly amazing. I am just starting with video and have been trying getting usable videos using Freepik. Since I need the best possible quality, I'm relying on Kling almost 100% of the time. However, since I'm using it through Freepik, a lot of features are not there. I've been struggling with a scene in particular and I must say, the tip you mention in another video about adding "subtle motion" has improved my results a lot! I'm trying to animate an image which is a front view of 5 people working in a modern office setting, with two people taking notes. I have made several attempts but every single generated video has distortions and I can't get even 5 seconds right. Like I said, adding "subtle motion" made it a lot better, even if it still didn't come out 100% right, so thank you! Is there any other thing I can try? At first I was using long and detailed prompts, now I'm trying short and concise ones like yours. I added "static camera, front view" as well, following your other tip about adding the angle to the prompt. This is a very difficult image to animate. Again, thank you so much for these videos. I'm learning a lot through them!
Hey, great to know you enjoying these guides! I haven't used freepik before, idk which version of the video model they are using. The more people you try to animate, the more likely there will be some distortions. I haven't seen any ai videos platforms that do a good job with that many people inside of them. Something like this might work "subtle motion, static camera, cinematic still, motionless"
@@taoprompts Thank you for your reply! As of now, they are using Kling 1.5 STD and Pro. I use Pro almost all the time. Thank you for telling me that, maybe I should keep the number of people in the videos lower. And thank you for the prompt tip as well, I'm going to try and see how it does! Hope you have a great week, merry Christmas and happy new year!
Hi I struggling to make income what kind of skills do I need to master to do ai work with clients to get into freelance work ? What are some pro tips please or a road map that you approach it with thanks and great video as always
nice video again. I´m a kling ai power user. generated 2.1k images last month with midj. and animate them with kling. got several ai tik tok channels. for my fusion channel i used the standard mode because the pro mode didnt get good results as yours. has kling improved in that?
I don't think they've made any updates to the model. The start and end frames is only available in version 1.0, so it's still the old model. It took 2-3 tries to get a good video for me, it could just be that the specific examples I used worked well.
I tried to turn a picture of a Victoria pigeon into a intricate 3D design of a bird in Kling and it didn't work.. I get from your video that I need an intermediate stade created in Midjourney, so im gonna do that first. Thank you again! 😉✨✨✨✨✨✨✨🌹
I've seen it work in Kling or Luma, make sure you set the image reference as the end frame. here is the original twitter post I mentioned from Halim: x.com/HalimAlrasihi/status/1835662381344588184
Tip 2 - Animal Hybrid Fusion Prompt ( @officialmrlemon ):
"zebra and a car leap into each other, colliding into each other. As they clash, they merge into a hybrid beast"
Tip 4 Facial Expression Prompts:
www.reddit.com/r/aivideo/comments/1g6shqk/minimax_expression_tutorial_imagetovideo/
Joy - Wide eyes, big smile, raised eyebrows, relaxed facial muscles.
Sadness - Downturned mouth, furrowed brows, teary eyes, drooping eyelids.
Surprise - Wide eyes, raised eyebrows, open mouth, slight head tilt.
Anger - Furrowed brows, clenched jaw, flared nostrils, narrowed eyes.
Fear - Wide eyes, slightly open mouth, tense brows, rigid jaw.
Disgust - Scrunched nose, raised upper lip, furrowed brows, narrowed eyes.
Confusion - One eyebrow raised, slightly open mouth, furrowed brows, head tilt.
Contempt - One side of the mouth curled up, narrowed eyes, head slightly turned.
Determination - Tightened lips, narrowed eyes, focused eyebrows, set jaw.
Shock - Frozen expression with wide-open eyes, slackened jaw, raised eyebrows.
Relief - Closed eyes, small smile, relaxed eyebrows, exhale shown with slightly open mouth.
Embarrassment - Eyes looking downward, blushing cheeks, sheepish smile, raised shoulders.
Tip 7 - Consistent Character Tutorial (FLUX):
th-cam.com/video/5Z8fwEeWfRg/w-d-xo.html
Tip 10 - Ink Illustration Prompt (@HalimAlrasihi):
"Revealing an image from complete blank paper transitioning into the style of ultra-fluid seamless smooth ink brushstrokes that dually and effortlessly flow, merging like water on rice paper. Strokes appear soft and blend naturally, deepening in boldness with perfect consistency. The image forms organically, with each element revealing itself in a continuous, graceful transition."
Full Tutorial: th-cam.com/video/XlT0vYGp1B8/w-d-xo.html
Thank you so much for the teachings! Very supportive and informative!... Now I just need perseverance and patience, right? 😂
@@mariepaulecoutens 👍🏾👍🏾 Definitely, and it's worth it. There are lots of tricks and techniques to be discovered by experimenting and creating with A.I. tools. Best Wishes
@FilmSpook Thank you so much for your support! 🕊️🌿⭐✨💫🍀💫✨⭐✨😉
I have problem getting ai to do wide angle shot showing a full body.. any idea of a better prompt ??
Thanks, Bro! In MiniMax there is a trick I discovered for getting a consistent character by reuploading a still frame from a generated video with that character. How it works is that you upload that frame in image to video and prompt something like, "Fast blur transition to a new scene showing..." and then you describe that character in a new scene or angle, or even with new clothes. MiniMax then tries to recreate that character in a new scene. It may take several or more tries, but when you get another scene that looks perfect with the same character, you then download that video and use a single frame from it to then upload back into MiniMax to create a full 6-second clip with the character in that new scene. However, sometimes I have used a video that has had the prompt for "Fast blur transition" and simply spliced and cut off the first second or two when in my editing software, and then used the remaining 3 or 4 seconds (that show the character in a new scene) in my short film. This method CAN be a little time consuming, I admit, hehe, because sometimes it takes two dozen generations to get the perfect facial resemblance of the character in a new scene or angle or with completely different clothing (if that is prompted). I was going to make a video about this technique, but decided to share it with you here, but I may do a video on it next month. Cheers, Brother. Carlos
Thanks for sharing!
@@OrigensSurpreendentes-j3c Very Welcome! 😎Cheers
That's a super creative tip man! I've noticed Minimax sometimes pans to a new version of the same character, but never thought about using that as a way of getting consistent characters
@@taoprompts 👍🏾👍🏾 Yeah, it really works well sometimes, with patience. Appreciate you, Bro
Yo brother thank u for sharing your experience pls make a video on it ❤❤❤
Hi, you are one of the best TH-camrs, and your work is amazing. I would like to say thank you for always sharing your work with us.
Thank you so much! Ai video has been improving so fast lately, I'm happy to share what I'm learning
Great teacher!! I think this video is like a gift to me~!! I haven't watched the video yet, but I will definitely apply for it after work today. Thank you :)😍😍
I tried to put in a lot of good tips in here, I hope you'll find it helpful!
The quality of these tips is insanely great. Didnt expect every single tip to be so actually helpful. Thanks for this!
Thank you! Ai video has improved so much lately
You are so underrated, only 17k views for such amazing informations 😮
I learn a lot from your videos, thanks for taking time to keep up us informed with the latest AI tools and techniques.
Of course! I'll keep you updated 👍
As always, excellent content again!
Thanks! I hope you found some of these tips helpful
AMAZING! Thank you Tao!!
You're the best in this. I always watch your videos when i wanna learn about AI ❤
Great to hear from you Jennie! There's some big updates coming up soon 👍
You made another great tutorial that solved my contrast issue in KLING.
Great to hear that! It's kind of frustrating when the contrast is always changing with Ai videos
Your video is always awesome!
Thank you, for your prompts😊
Your video is very informative for me I m watching now
I combined as many tips & tricks as possible!
I absolutely LOVE Kling!! The price is reasonable, given all that it can do and the movements and details are phenomenal. I can't even imagine what Version 2.0 will bring.
I'm excited for version 2.0 as well! Although I hear they even have a consistent character video feature planned soon
Super useful, thanks so much dude!
Subscribed and notifications on. nice work man
Thanks for subbing!
Thank you for your informative videos! You are my favorite author ❤❤
Thanks I really appreciate that! And I'll keep you up to date 👍
Lovely video, thanks ❤❤❤
Thanks for the love!
Thanks for this post makes a workflow using a video so easy to follow.
Thanks! I want to make more tips & tricks guides like this one 👍
Very informative & useful with me. Thank you so lot!
Glad this guide was helpful!
Thank you sir for this video
As always your video is very informative, thanks
Glad to be of service!
Awesome and educational video as always
Thanks Sky! I'm always looking for new Ai tips
RESPECT YOU,SIR
You are the best in this.
Thanks! I appreciate that bro 🙏
Really good tutorial.
Thank you ☺️
Glad I can help man!
This is one best AI secrets for a beginner like me. Am a Kling user. Thanks brother :) Cheers
Glad this helps man!
great as always 💥Thanks
Thanks Sean!
great content bro, so many sell outs in this space, you gonna be big, keep it up!
Thanks man! I appreciate the support 🙏
thanks, very helpfull video
We appreciate your kindness 🙏
I appreciate your support!
So we're "tweening" video now? That's a Flash term. What a trip. Thanks Tao! GotoAndPlay, bro. LOL
There's a lot of great video platforms so I wanted to put the tips together 👍
Thanks Tao!
great tutorial!
Thanks Abdul!
the voice and his look.. sounds like I've seen him in a movie.
Dude that was so helpful
Glad to help man!
This is helpful😊
Glad this helped you out!
Hi Bro can please make a video for probe lense shots
Video prompts
Thank U! To explain well! I asked a question from a former video, so the solution is to describe it specifically. So is it difficult to make character in AI tools?
I answered your other comment!
Tao - THANK YOU for all the great teaching. I am learning so much. I am struggling with one challenge and that is interacting hand movements. Using Kling with motion brush and prompts I just can't seem to get smooth natural movements. Also the hand placements always seem to miss where I show the motion going. For example placing a hand on an arm (bicep area) and it consistently goes to the chest or shoulder. In addition the skin as it is touched the AI makes it look leather like and not real. Any advice or perhaps a video tutorial would be helpful. BTW using Chatgpt has been a major help. Thanks again for all your teachings!
Hand motions are super hard to get correctly, especially between multiple people. I'd recommend trying a closeup shot on just the hand touching the bicep. So start with a reference image of a bicep and try animating a hand touching it
@@taoprompts Wow if you say it's "super hard" that says a lot! But I will try your suggestion. Thank you Tao
@@taoprompts So I tried your suggestion on Kling, Minimax, Pixverse & Luma. Kling was the only platform that handled it perfectly. But when I then included the full arm Kling, Minimax, Pixverse, Luma all missed it and went to the forearm ??? I even tried adding comments in the Negative Prompt input. And I even tried using Chatgpt to assist by showing it the image, prompt structure and generating prompts to try and each time the generation keeps goes to the forearm. Man I am so frustrated!! It would seem to be a rather simple task request why so difficult!?
I tried your suggestion and Kling AI worked perfectly, other platforms failed. However when I expanded the image to include the full arm Kling and others still placed the hand on the wrong location. I even tried Chatgpt using the prompt formats of each platform and still no success. I can't figure out why the models struggle with such a simple action? Any thoughts?
Hello there, I discovered your channel and I think your content is truly amazing. I am just starting with video and have been trying getting usable videos using Freepik. Since I need the best possible quality, I'm relying on Kling almost 100% of the time. However, since I'm using it through Freepik, a lot of features are not there. I've been struggling with a scene in particular and I must say, the tip you mention in another video about adding "subtle motion" has improved my results a lot! I'm trying to animate an image which is a front view of 5 people working in a modern office setting, with two people taking notes. I have made several attempts but every single generated video has distortions and I can't get even 5 seconds right. Like I said, adding "subtle motion" made it a lot better, even if it still didn't come out 100% right, so thank you! Is there any other thing I can try? At first I was using long and detailed prompts, now I'm trying short and concise ones like yours. I added "static camera, front view" as well, following your other tip about adding the angle to the prompt. This is a very difficult image to animate. Again, thank you so much for these videos. I'm learning a lot through them!
Hey, great to know you enjoying these guides!
I haven't used freepik before, idk which version of the video model they are using.
The more people you try to animate, the more likely there will be some distortions. I haven't seen any ai videos platforms that do a good job with that many people inside of them.
Something like this might work "subtle motion, static camera, cinematic still, motionless"
@@taoprompts Thank you for your reply! As of now, they are using Kling 1.5 STD and Pro. I use Pro almost all the time.
Thank you for telling me that, maybe I should keep the number of people in the videos lower. And thank you for the prompt tip as well, I'm going to try and see how it does! Hope you have a great week, merry Christmas and happy new year!
retro future please =D
I can see you like retro style videos from your channel 👍
Yo bro how to create a prompt with tiger and the tank hybrid?
how did you get the photos each of the e guy and then with the dog head? something with inpaint?
I showed how to do it in Midjourney after the video part
Can you please create a prompt for car image to video generation
You can just describe what the car is doing, ie "driving forwards"
@taoprompts while giving forward prompt in kling ai sometimes the car goes reverse
How can i make The people on videos to stay still?
Hi I struggling to make income what kind of skills do I need to master to do ai work with clients to get into freelance work ? What are some pro tips please or a road map that you approach it with thanks and great video as always
Hey man, you can always message me on instagram @taoprompts if you want to chat
nice video again. I´m a kling ai power user. generated 2.1k images last month with midj. and animate them with kling. got several ai tik tok channels. for my fusion channel i used the standard mode because the pro mode didnt get good results as yours. has kling improved in that?
I don't think they've made any updates to the model. The start and end frames is only available in version 1.0, so it's still the old model. It took 2-3 tries to get a good video for me, it could just be that the specific examples I used worked well.
Pls, can you recommend a free image upscaler to produce the best result for videos in minimax.
Thanks
Upscayl is free to use and works well
Thank you
I tried to turn a picture of a Victoria pigeon into a intricate 3D design of a bird in Kling and it didn't work.. I get from your video that I need an intermediate stade created in Midjourney, so im gonna do that first. Thank you again! 😉✨✨✨✨✨✨✨🌹
If you can add an intermediate state that will work the best! Also it's also worth trying Luma to see if you can get better results there.
@taoprompts Hooo! OMG😅Thank you for your advice! Your comment section is so nice and helpful... 😉
the brush illustration doesn't work both in klain and luma, i don't know if it's me or the engine
I've seen it work in Kling or Luma, make sure you set the image reference as the end frame.
here is the original twitter post I mentioned from Halim: x.com/HalimAlrasihi/status/1835662381344588184
I dont get it how can i put my video there and add something on it using AI based on prompt? Or is it impossible?
No you can't use your own videos right now
Brother, just a query. the Dog-Man picture. May I know how did you create three different heads(dog,man,oldman) for the same body??
Hey, I showed how to do that in this video using Midjourney image editing
There is no such thing as "consistent characters" in text-to-image or text-to-video AI. That's what image-to-video is for.
image-to-video is the best way of doing it
can someone finally tell me if a tiktok vid has to be 1min long? everyone is saying something different