New modeling technology breathes life into animation
By Darren Murph 
posted August 19th 2008 10:25AM
posted August 19th 2008 10:25AM
[Thanks, Przemek]
posted August 19th 2008 10:25AM
A look back on popular stories from today in a specific year.
Now that we've thrown 'em off the trail, use the form below to get in touch with the people at Engadget. Please fill in all of the required fields because they're required.
OMG!! somebody is milking it if it takes till 2020! That is just too freakin creepy.
It looks great, but they're not quite out of the valley yet. The eyes and mouth just don't look right in certain places. And the biggest problem is that this is just another form of capture technology, not a new animation technique. If you have to take a video of someone, just to recreate the exact same video, what's the point? When we get this level of quality on original animation, then it will be impressive.
Ok I've been working in the 3D animation industry for a long time, this is a special type of animation called motion capture. Her face looks really realistic and great but I was not so impressed to read that everything else is real and they just stuck it on top. Mo-cap has been in use for a long time mostly to avoid the uncanny valley. The valley poses much more of a threat to those animating by hand (with keyframes and curves). Most animators can't afford mo-cap technology and probably won't be able to mo-cap everything they need to animate so it's not like this is the dawn of some new era in animation.
The CGI face is way more attractive than the human face in that video though.
This video is very impressive!
However, the problem of the uncanny valley goes much further then animation or symmetrical/asymmetrical modeling. Symmetrical modeling has pretty much been wiped out since programs like zbrush or mud and I dont know of too many animators that would rig only half of a face animate and then mirror for a project that revolves around a bust framed short clip. The uncanny valley shows in any bit of visual information presented on screen that we subconsciously pick up as not making since. Thats why the uncanny valley is also referred to in the prosthetics industry. When is the last time you saw and amputee with a realistically colored AND shaped prosthetic? This is often avoided today by making the prosthetic any color but flesh tone or by avoiding looking like the body part it is to replace.
It seems as though when we see something so OBVIOUSLY not real, it looses the uncanny valley.
This is also why we often think Pixar movies look so amazing as we have already achieved a suspension of disbelief through non realistic characters. This allows us to look at the graphics and animation without trying to pick it apart visually.
The tragedy is that the animation indeed looks very real but.. the fake mascara really ruins it!
Now I can put my face on a pron film, and put the face of my crush on the girl.
Due to memory constraints, you will have to use your face for both actors. We hope you enjoy your customized video! KTHXBYE
It is NOT animation but just a newer motion capture with added pixel data process. Basically putting multi camera pixel data on polygons.
It can not do anything different ( well a little tweaking and allow the captured data to be seen from all angles) than what the actor did in the original shoot. So not animation which is defined as "to give life to". It is not a creative, imaginative artistic process which is at the heart of great animation, great CG movies and great games. It is just a brute force transfer technique that Hollywood types like because they simple hire an actor - rather than having to deal with real talent of animators and art directors. And of course it could not do anything a hired actor can't do ( which is the whole point of SF/CG movies and games to go beyond to new realms), so no creatures, animals, mythological characters, godzillas, yodas ....
Too bad you and all the other sites fell for and just repeated the company's lame ( an untruthful) marketing hype.
Andy Engelkemier (above) writes correctly: "A good example of the strangeness would actually be reanimating familiar, but dead, actors. Imagine what they would look like with someone else's facial mannerisms. It's creepy as hell. It's worse when we know what they should look like, but it still feel awkward when we are looking at an animated character we don't know when it is trying to be realistic."
The rest of you are surprisingly gullible for engadget readers! This is essentially just a very high-quality rendering of that particular person. The data they've assembled on her would have virtually no value if applied to either another face and body or an original design. This is largely a scam. Grow up people.
I want to shoot her in a game... However disturbing it may be.
It's actually very impressive! Her eye blinking looks unnatural to me. And as someone mentioned "subtle" movements are bit over-exaggerated
I don't see how this is 'new' or particulary clever, of course you can break down everything in smaller and smaller parts, both movement and texture, but that doesn't mean it's practical in real use, you can have a supercomputer run a raytrace sequence of super high definition and radiosity etcetera, but that doesn't mean you can use it in games any time soon.
If you look at movies you already see that they use rendering and animation with a much higher than standard set of samples, but games are going through their own paces, and games mostly benefit from what ends in public domain anyway, and what has tutorials so they all start to use it, and of course it must be able to run realtime.
Emily's pussy can talk?
Oh, POSSE.
It's seems amazing at first but once you realize only her face is generated, it's just ok. The mouth seems a bit off too don't you think? Ok it's probably the best yet but having the ACTUAL real footage filling most of the frame fools you somewhat.
It's just her face that is animated...her body and hair and the background are all actual video, used from the original plate....I guess it's still interesting to see...but I look at the hands when viewing animation...hands look more fake to me than the eyes...
I cant wait for Emily and Stasha to be doing some hot lesbo porn
Well, however they manage to make an animation look like a human, this video seems to show they're getting better and closing the gap, but the mouth doesn't move correctly, and the eyes are closer to the finish line than the mouth, but they too, don't move *exactly* right - her eyes don't reflect enough life or something; it's like they're too "flat". The head and neck, arm and hand movements seem to be right on, though.
I don't know exactly how to describe it, but the mouth just doesn't move correctly - the teeth don't look right, and her lips aren't relaxed enough when she forms words and so forth. Her mouth also moves a bit too slow, as well.
They've never been able to reflect human likeness and all the incredibly subtle nuances of the human face. I highly doubt we're going to be waiting until 2020, though, to see it - I don't know where the CTO from AMD gets that year from. I believe we'll get there much sooner than that.
What I'm wondering about, is when they'll be able to create a model without any video mapping of a real person's face. Now that would take a real and very deep understanding of the human face.
It looks like they used a 3D model and fed motion-capture data into it. Motion capture data will add in the nuances you need to make things look more realistic, but mocap is not 100% simulation.
The uncanny valley is actually just below the face; get to work people.
That's not the uncanny valley...that's C*** valley. But I understand your misunderstanding.
This article is titled: "New modeling technology breathes life into animation".
No. Not true. Animators breathe life into animation, not tools. Pencils don't make the animation happen in hand-drawn animation
and computers don't do the animation in CG animation. Tools (whether pencils or computers) are used by artists to breathe life into the animation.
That said, the clip that is being touted as something new and improved in terms of photo-realistic CG still has that
creepy-zombie look to me. None for me thanks.
The highest goal of Art is not literal-minded photo realism, but to caricature real life. The creative artist interprets and entertains . The non-creative technician only cares about how many hairs and pores can be programmed into a "realistic" character. Technology increases , but a sense of good taste and aesthetics doesn't necessarily go along with increased technological ability.
You are right new modeling technology breathes life into animation. Today lots of software are available and with the help of them one can create superb 3d models that looks like alive.
AutoCAD Drafting Services