"Head Sliding" vs "Head Pointing"

[wsmith]

I've been an off-and-on SmartNAV user for a few months now and am getting a bit frustrated. Let's clarify the two types of "head-tracking" commonly used today:

The SmartNAV implements a two-degree-of-freedom mechanism; internally, the camera merely translates the dot position into a 2-dimensional x,y coordinate. This means that the device does not detect "head pointing", i.e. the direction you are looking, although it may seem that way initially. For example, do the Dana Carvey "Turtle, turle" move [slide your head up and down without tilting it] and you will see what I mean. While using the product, you cannot move your head without moving the mouse. This commonly results in stiff necks in the initial "burn in" phase. Eventually, your neck muscles become accustomed to staying still and it feels natural. Essentially, you are creating "muscle memory" [or constant spasms] in your neck, which will eventually cause RSI-like symptoms in your neck.

The TrackIR and all newer products use a true head-pointing mechanism with more degrees of freedom. They translate your head into a three-dimensional vector and can "see" what you are looking at. This is usually accomplished using multiple dots [see the Johny Lee's Wii-based head tracker for example].

The SmartNAV costs considerably more than the TrackIR, so when I purchased it, I was expecting a better product. I was also expecting the neck spasms to go away. I feel that this is not the case however. The new TrackIR has the same specs as the SmartNav 4, but with the mouse disabled.

When will we, the SmartNAV users receive a product comparable to what we paid for? When will the SmartNAV support more DOF to reduce neck spasms? This technology already exists in the TrackIR. What gives? I can't wear both devices

-Warren

[Seth Steiling]

Hi Warren,

A few important things to consider when comparing TrackIR and SmartNav:

- From a hardware standpoint, SmartNav definitely is a superior camera. It offers more precise tracking and better illumination. The products use different image sensors, and the one in SmartNav is superior in pretty much every way. Even though some of the specs appear similar on the surface, what's going on beneath the surface isn't comparable (different shutter type/speed, different image processing, different lensing, more durable metal case, strobed illumination for enhanced background interference removal, etc). That's where a good portion of the difference in cost comes in.

- SmartNav is designed first and foremost for the Assistive Technologies market. So, it's very important for the design to be as minimalistic and simple to use as possible. For the majority of our SmartNav users, a hat or headset as a required accessory is not feasible. They need to be able to use the device for hours at a time without needing to adjust a hat, hit a recenter button, etc. Now, these features (hat, recenter, etc) are still available to SmartNav users, but they aren't required. If we used Vector tracking instead of dot tracking, it would complicate the usability for the majority of our core users.

- We have extensively tested Vector tracking for 2 DOF mouse control, and we determined that it does not offer a significant improvement of tracking accuracy. The benefit of Vector tracking is in the immersion it offers in 6 DOF applications. But from the standpoint of raw tracking accuracy, it does not offer an appreciable improvement to 2 DOF dot tracking. In part, this is because with Vector tracking, accuracy isn't only determined by the resolution of the camera. You have to apply additional math to the marker tracking, which can introduce more variable results. Now, we've hammered on our math significantly over the years with the TrackIR line, and it has yielded pretty exceptional results. But it's all relative. Single marker tracking, when you're only using it for 2 DOF mouse control, still offers a mathematical simplicity and reliability that is very optimal for the SmartNav product.