Electromyography and Headtracking

Wow, amazing is right! I think these folks may very well be on to something. The idea of tracking movement with muscles does seem alot more intuitive than using cameras, magnets, etc..

If it comes close to what's in the video, that's pretty damn cool.
I've never even heard of this device and they already have 25000 preorders for the first batch. Second batch isn't going to ship until 2014.
After surviving the Pandora I'm a bit hesitant to order something I don't know much about that will take a year to deliver.

From their faq:

What sort of precision does the MYO have?
The MYO detects gestures and movements in two ways: 1) muscle activity, and 2) motion sensing. When sensing the muscle movements of the user, the MYO can detect changes down to each individual finger. When tracking the position of the arm and hand, the MYO can detect subtle movements and rotations in all directions!

That makes it sound like it has accelerometers to track motion. The rest of the site is a bit vague on the tech.
It can detect each finger, but can it tell how much each finger is curled?

It wouldn't really work well for head tracking, it would only give crude neck orientation which we can already do better. It might, however, be great for hand tracking, depending on how accurately it reports finger position. An armband would be a lot more acceptable and comfortable to the average consumer than a glove. Anybody ever use the P5 glove? it was pretty accurate but it was actually painful to wear.

There was an experiment in the '80, The Libet experiment, which broke loose a huge discussion on free will.
"Libet found that the unconscious brain activity leading up to the conscious decision by the subject to flick his or her wrist began approximately half a second before the subject consciously felt that she had decided to move."

So here is the thing: to give the mind the illusion of exact concurrence the signal had to start its way down the nervous system before it is consciously. And they proved that this is what happens: the signal is send before the mind is aware what it wants to do.

So that's kinda the tech here, you don't have muscles in in your fingers - your fingers are moved by pulling strings in your underarm and because it takes a little time from pulling that string to actually move your finger the detection could appear faster. Sense for your self, grab your underarm tightly and make a fist (try the same with your neck and move your head).
Or think of an real world martial arts fight, when you wait for your opponents attack move to counter it, you do not watch for his hands or arms to move. What you want to do is to focus on his shoulder movement.

Because the myo is worn below the elbow, they might need the motion sensor on top for the big picture of upper arm and shoulder movement combined. ERP for finger tracking and the motion sensor for the world position of the arm.

it's a neat idea, but it seem like for it to be useful in hand tracking it would require another technology. Like if you could combine it with the magnet positioning of the razor naga hydra. But like other said, there the question of it can detect how much each finger is open.

Tracking the neck is great; but what happens if some one is turning with another part of their body.

Just thought I'd note that this tech could have more potential than we think. Its use in prosthesis gives me hope that this could possibly be used full on as a control system for using the arms, legs and neck to manipulate in game character models to a whole new level. I'll be going in to the concept in the near future in another thread, but let's just say for now that the tech could give us enthusiasts a taste of VR's true potential.