Hello Marc! May I say how it works? Stabilize the video, then use any free AI tool to generate a Z-Buffer depth per frame, then use the video plus the fake depth to trick Metahuman pipeline to think the video comes from an iPhone. As long as it sees a depth map and a video it will track it and animate it. How do you convert 2D video to stereo to get the Depth map per frame? Google it, but there is one called StereoPhoto Maker. It can be done in Photoshop using the Depth Blur neural filter and you could automate a batch processing using actions.
hi , the only way i know and i show , its use keen tools to track face and use same tracked model face to get the depth , blender have keen tools , and nuke, but could use whatever software to do same
Hey, Marc. This is the youtube game, bro. You could indeed make those presentation improvements so your message gets spread more, since you indeed have LOTS to show. We need voices and experience like yours, dude. Some things from this video may look obvious to you, because you have lots of experience, but for others, it is harder. I didn't get how did you add custom depth for instance. But okay, I'm glad at least you share this bit. Thanks, man!
Your Metahuman face rig transfer is the best I’ve seen, great range of different face shapes! Is the typology different for each model or all based of the Metahuman typology?
a bit late but can say base rig is all based in metahuman from there as i show in my last custom topology video , can transfer weights and blendshapes to a custom topology
thanks but i gave 99% of the hint .... main point to achieve this its track face with software like keen tools to make it stable like if the actor in the movie had iphone in his face..... the last hint would be say everything and even invite u all to coffee and thank you for your support... my videos have 5-10 likes average like nobody very interested......i show main part of the process to give ideas , at least dont push me to say where is the X to find the treasure with 0 effort, i dont push anyone to click like and can spread more the videos and with more support could think about share some more stuff ......... im not very into social media and im not into play the social media game to try get more attention... but after all im always open for fast straight forward freelance stuff
@@marc1137 thanks Marc, I'm able to figure the "video" part, but I'm stuck at finding solution to the "custom depth" you showed here. anyways your works are hidden gems, I'm glad to find your works here.
@@afkCG The answer is. Use AI to generate depth maps per frame, the metahuman pipeline is expecting video and depth information from the stereo cameras/iPhone depth camera. Use Photoshop neural filters, depth blur.
just need someone else with iphone do a livelink video and send you those folders with the basic data , then replace image sequences for your new sequence and start to see what happen and how fix things , but main point is u dont need an iphone
I think it’s incredibly selfish in this industry to not share your findings and keep things a secret just because you have “very few likes” as you put it. The whole spirit of games design, visual effects and anything creative is to share new techniques and people iterate on these new discoveries so new people can create really cool things! Please consider sharing how you did this, it doesn’t have to be a step by step video but even just saying the overall full process rather than hiding how you managed to get the custom depth which is the main thing people want to know. You gain nothing by keeping your methods secret but if you share with the community you gain the respect of many people as well as the great feeling of helping other people.
you twisting the story of the few likes and putting that whole "community" in your side to attack me with the romantic idea of share whatever one discover ......well i shared last weeks a nice wrap tool idea , seems if its not useful for u , its nothing.... so im selfish not sharing the last trick of the thing that would be useful for u ... and i say last trick because this animator stuff i shared 99% of the process showing what im doing , for now i just keep for me the last silly trick and letting people explore themselves, and i dont care the few likes , and one of the points is i dont want go deep into this social media jungle , i post stuff people can feel inspired and interested and shared a lot of things, and as i said im not into social media so i wont keep replying and start that kind of show , you said clear ur point , i replied clear my point , and for me is incredible theres a 2% of people commenting videos in this ways... i also was in the other side watching videos about someone showing something i would like to know and not saying how , and i never ever would dare to comment that video saying u are incredibly selfish in the name of the whole community
@@marc1137 I was simply stating my opinions and never asked anyone to attack you over anything, that would be completely immoral. I just fail to see your reasoning behind not sharing your knowledge is all.
Honestly its not difficult at all, I made my own solution that uses ai tor the depth and it results in better animations that using phones, you can also use mediapipe of you want a solution without ai that's runs faster as using AI uses alot of vram
i tried getting solved metahuman from video footage (not livelink app with depth) and of course it gave me a metahuman with a flat face 🤣. how did you make this?
i think in the video im showing 90% what im doing , i wont do a step by step 100% tutorial.. when i feel not much interest with my 15 likes average videos , its funny all those meta influencers showing easy simple stuff with a cool microphone and thousands people behind , better stop try find step by step - think nothing stuff , and see what im showing which is the 90% of the thing and try be curious and find some soution by urselves.. really is good life experience...... than just go to the places and ask ... how u do it , thanks...... im saying all this just to help people , not pissed off or something 😅
Hello Marc! May I say how it works? Stabilize the video, then use any free AI tool to generate a Z-Buffer depth per frame, then use the video plus the fake depth to trick Metahuman pipeline to think the video comes from an iPhone. As long as it sees a depth map and a video it will track it and animate it. How do you convert 2D video to stereo to get the Depth map per frame? Google it, but there is one called StereoPhoto Maker. It can be done in Photoshop using the Depth Blur neural filter and you could automate a batch processing using actions.
hi , the only way i know and i show , its use keen tools to track face and use same tracked model face to get the depth , blender have keen tools , and nuke, but could use whatever software to do same
The way you explained, I tried many times. UE won't buzz. It quits the process right away. Whatever Marc explained in response is the way to go.
Hey, Marc. This is the youtube game, bro. You could indeed make those presentation improvements so your message gets spread more, since you indeed have LOTS to show. We need voices and experience like yours, dude. Some things from this video may look obvious to you, because you have lots of experience, but for others, it is harder. I didn't get how did you add custom depth for instance. But okay, I'm glad at least you share this bit. Thanks, man!
Hello! Nice work I have one question, how you transfered Horse to metahuman?
Your Metahuman face rig transfer is the best I’ve seen, great range of different face shapes! Is the typology different for each model or all based of the Metahuman typology?
a bit late but can say base rig is all based in metahuman from there as i show in my last custom topology video , can transfer weights and blendshapes to a custom topology
You’re awesome this is very helpful
inawe bro, keep up
This is really cool, your work is truly inspiring, thank you for sharing that it can be done, please at least give us a hint.
thanks but i gave 99% of the hint .... main point to achieve this its track face with software like keen tools to make it stable like if the actor in the movie had iphone in his face..... the last hint would be say everything and even invite u all to coffee and thank you for your support... my videos have 5-10 likes average like nobody very interested......i show main part of the process to give ideas , at least dont push me to say where is the X to find the treasure with 0 effort, i dont push anyone to click like and can spread more the videos and with more support could think about share some more stuff ......... im not very into social media and im not into play the social media game to try get more attention... but after all im always open for fast straight forward freelance stuff
@@marc1137 thanks Marc, I'm able to figure the "video" part, but I'm stuck at finding solution to the "custom depth" you showed here.
anyways your works are hidden gems, I'm glad to find your works here.
@@afkCG The answer is. Use AI to generate depth maps per frame, the metahuman pipeline is expecting video and depth information from the stereo cameras/iPhone depth camera. Use Photoshop neural filters, depth blur.
Amazing effect! I also think there is no shortcut to success, but I don't have an iphone, can I make it like you do? Or do I have to have an iphone.
just need someone else with iphone do a livelink video and send you those folders with the basic data , then replace image sequences for your new sequence and start to see what happen and how fix things , but main point is u dont need an iphone
This is nuts!!!
thanks, valuable comment coming from experienced metahuman tester 🙂
i like it your animal facal animation!!!
yes its interesting also can use it there
Legit work mate, touché
I think it’s incredibly selfish in this industry to not share your findings and keep things a secret just because you have “very few likes” as you put it. The whole spirit of games design, visual effects and anything creative is to share new techniques and people iterate on these new discoveries so new people can create really cool things! Please consider sharing how you did this, it doesn’t have to be a step by step video but even just saying the overall full process rather than hiding how you managed to get the custom depth which is the main thing people want to know. You gain nothing by keeping your methods secret but if you share with the community you gain the respect of many people as well as the great feeling of helping other people.
you twisting the story of the few likes and putting that whole "community" in your side to attack me with the romantic idea of share whatever one discover ......well i shared last weeks a nice wrap tool idea , seems if its not useful for u , its nothing.... so im selfish not sharing the last trick of the thing that would be useful for u ... and i say last trick because this animator stuff i shared 99% of the process showing what im doing , for now i just keep for me the last silly trick and letting people explore themselves, and i dont care the few likes , and one of the points is i dont want go deep into this social media jungle , i post stuff people can feel inspired and interested and shared a lot of things, and as i said im not into social media so i wont keep replying and start that kind of show , you said clear ur point , i replied clear my point , and for me is incredible theres a 2% of people commenting videos in this ways... i also was in the other side watching videos about someone showing something i would like to know and not saying how , and i never ever would dare to comment that video saying u are incredibly selfish in the name of the whole community
@@marc1137 I was simply stating my opinions and never asked anyone to attack you over anything, that would be completely immoral. I just fail to see your reasoning behind not sharing your knowledge is all.
Honestly its not difficult at all, I made my own solution that uses ai tor the depth and it results in better animations that using phones, you can also use mediapipe of you want a solution without ai that's runs faster as using AI uses alot of vram
Great Video,But i almost there,no need to chase it anymore
congratulations and well done
i tried getting solved metahuman from video footage (not livelink app with depth) and of course it gave me a metahuman with a flat face 🤣. how did you make this?
i think in the video im showing 90% what im doing , i wont do a step by step 100% tutorial.. when i feel not much interest with my 15 likes average videos , its funny all those meta influencers showing easy simple stuff with a cool microphone and thousands people behind , better stop try find step by step - think nothing stuff , and see what im showing which is the 90% of the thing and try be curious and find some soution by urselves.. really is good life experience...... than just go to the places and ask ... how u do it , thanks......
im saying all this just to help people , not pissed off or something 😅
How did you map a horse face to metahuman?
moving joints and re skin... some previous videos show part of the process , u can check metapipe also
not gonna lie.....i understood nothing
How this works?
Nice work!
thanks