Project Natal drops hardware Motion detection to save costs!

[JaggedSac]

drkohler said:

WereKitten said:

JaggedSac said:

I don't think Natal can distinguish fingers anyway.  It can measure where the arm is pointing.  My question is whether it is a cube with volume of 4 cm?  Or is it a cube with 4 cm on each side?

If the x/y resolution is 4 cm, that's even worse and you have to cut that in more than half, with on-screen steps of an inch and a half.

It is pretty clear that the 4cm correspods to xy resolution. Let's assume that the tof chip is an _expensive_ one with a pixel count of 320*240. with those 320 pixels you have to cover max 4 players hopping in front of you - about 5-6m wide space/320 = 1.6-2cm per pixel. The z resolution could be better depending on whether the tof camera depth resolution is programmable or not by developpers. 4cm z resolution would indicate the depth space covered is 256*4cm = 10meters which seems a bit high.

Is Natal using TOF?  I read they are just reading the light intensity reflected back from the IR bursts.  I thought TOF based it on the time it took and not light intensity.  I am assuming they would correlate, but TOF is probably more expensive and accurate.