forget motion capture - think about Image Metrics

Check this - quite impressive

links are in Germany but movies are in Engl.


Making off

I keep seeing this. Good work. I’ll be happier when it’s cheaper:D

It’s simply amazing… The first time I saw this kind of capture I got stunned. Imagine all the possibilities… How much quality artistical expressiveness will be possible for low-budget/few-people studios (when it becomes affordable), reducing production times and being able to achieve much bigger productions done by fewer people.

I know Mocap techniques sometimes are seen as less prestigious than manual animation, well simply because the artists didn’t do it. But I believe it depends on the goals of the project. If the intention is to tell a story, express ideas, transmit feelings, and these tools do the job, well, that production will not be awarded for the skills of the animators, but the goal of telling a story, expressing those ideas/situations/feelings would be met, with less work and in a shorter time. And this technique allows also the mix of both captured and hand made retouches. So it’s trully amazing.

some other vids:


This is the first time I’ve looked at the the higher res clip. I have to say, that just looks scary.

It has that feeling that something is off but you can’t quite finger it. You know, that uncanny nonsense:cool:

If you ask me, it looks really nice but I wouldn t say that they are beyound the uncanny Valley, perhaps I m wrong but that is my opinion, I mean you still can see that this is CG. On the whole I think that it is interesting what they ll do next :slight_smile:

I was thinking myself “the face, especially eye and mouth area looks juuust a tad too bland somehow, but the hair is perfect”. Then I realized that the face is only thing that is animated. Looks very good anyway.


I’ve seen some image metrics stuff before, but they had me for at least the first half… i already started wondering why they’d hire someone to explain the tech, instead of showing it… must be an actress, not an intern, … then a spark, duh, it’s rendered, they’ll show an overview in a moment… !

so, from now on, anything said on video / televison can be fake !

This is quite nice. Is it purely derived from image processing or are they using a z camera to capture z-depth?

Are they using data from multiple cameras or can it track from just one viewpoint?

Michael, afaik it’s a normal single camera. I think it’s fixed in front of the face.

Looking at the animated mesh, not the face, it seems like fairly simple adjustments and the wrinkles etc seem to be added in as a specular/normal map. Anyone else think this?