Read the article, it says clearly that "frame sequential display and active-shutter glass systems" are being used, i.e. 120Hz TVs and LCD shutter glasses, NOT polarised light.

Means more expensive glasses, but the 3D pictures are higher quality.

That's actually incorrect. Sony's (now announced) system uses "frame sequential display and active-shutter glass systems", i.e. 120Hz displays and LCD-shutter glasses like nVidia's 3D Vision, not polarised glasses.

Polarised glasses are cheaper, but can only be used with specially-polarised displays at half the resolution, or with dual-display systems. This is a better system for cinemas, but LCD shutter is probably a better system for the home.

Not necessarily. HDMI 1.4 is the first to officially support 3D, but all that's really needed is the ability to handle a 120Hz signal. According to the spec, HDMI 1.3 supports "beyond 120Hz", and even has enough bandwidth to do it at 1080p.

No, not polarised glasses, active LCD shutter glasses required (like nVidia's 3D Vision).

Shutter glasses are more expensive, bulkier, and have to be recharged - but usually have less ghosting & better quality, and more importantly they work with any 120Hz-capable display, like DLP, plasma and a couple of newer LCDs.

Good confirmation of the capacitive screen reported for the upcoming Leo.

I doubt the stylus is "magnetic" though, probably just conductive in the tip (like a finger).

Problem with glasses-free (lenticular) displays is that if your head is not in the right position, the effect gets messed up. This TV apparently can let about 4 people watch at the same time, but only if they're at just the right positions and angles. Also, you get half the horizontal resolution.

Personally, I don't see the problem with wearing glasses, since most people at least wear sunglasses occasionally anyway, and you generally get better quality. Passive-polarised displays are easier to view, though half the brightness, and need only ordinary sunglasses with differently-polarised lenses. Active-polarised displays give the best results but need bulkier LCD-shutter glasses, though these are getting rapidly smaller.

Ah, but was it a 400MPH airblast?

Did you *measure* it??

I actually get significantly better results (using CoreAVC) from CUDA than from DXVA. Dropped the CPU usage on my old AthlonX2 HTPC from 80-90% down to 10-15%, and it works on a wider variety of formats too, in my experience.

Face/voice recog is entirely software, not hardware - Microsoft's domain, not Sony's.

Natal hardware is just a webcam with a clever IR cam that senses depth. Everything else is software.

Trailer? Faked. On-stage demos? Not faked.

Read the reports from journos who've played with it first-hand. They're impressed. Natal does actually work, at least so far.