Also the middle model looks a tiny wiener bit gay the way he dances. But aside from that the right model looks so fucking human.

probably look at people with dance and performance backgrounds, (kinda like how wrestling games use trainers and trainees for motion capture performances for video games)

Step 1: record yourself dancing 5,000 times Step 2: watch recordings Step 3: ????? Step 4: profit!

well step 1: you have to be born. . . yeah that's as far as I know

Mostly has to do with lack of important details on the left model: no shadows, ankles never move, static facial expressions, fingers barely move, the model is not stuck to the ground. Makes me wonder if the problem with the model and not the mocap itself. They should've used the same model for better comparison. Besides, a good animator could clean up the issues (like they do anyway with mocap) and it will look very convincing for an untrained eye. Not good enough for titanic movie/game productions, but they would buy the expensive one anyway, and for a smaller studio or individual this is perfect price for quality

I don't think you're paying for the models. You're paying for how well it can sync your motion within the suit to the model you have. My understanding is that the $100 option is using the video to try and sync the motion to the model on the left. So it's sort of watching the video like us viewers, and basing the movements on that. But because of that, it doesn't have as much data to more accurately make the model move like the dancer. The model on the right though, is using the actual points on the suit, from the front and back. So the model is more likely to move like the dancer. As someone mentioned, the $100 option only has access to 2d space. So it's always going to look off, it has no frame of reference for 3d objects. I wonder if there's a way to sort of make that less of an issue though. To have the program still use video to sync up the animation, but somehow have it respect 3d constraints.

it specifically states markerless. you can do proper mocap for free with right software.the uploader picked markerless because its the cheapest & most janky whereas optical is more expressive & non conventional meaning it cost more but neither are the best mocap techniques.its just for views

@@tiacool7978Rokoko studio has a 20$ a month streaming mocap package that supports 2 webcams, meaning you can gain extra accuracy adding another perspective angle.

No. What you would want to do is a facial cap with a different rig. You could add tracking to it but it would have to be fixed from here to next week because of all the movement artifacts it would cause.

One having the capability to track and represent real time location is certainly part of the comparison and does not invalidate it.

It's fair enough when you factor in the original Optitrack one was rendered by the original dancer/producer. The markerless one was effectively added on top, probably only using the original video as reference. Not even sure they asked for permission from the original owners.