Your points are well taken, Ian. However, my interest lies in the realm
of telepresence, where maximizing FOV improves immersion. Any person who
has worn a commercial HMD with even 60 degree displays can attest to the
sense of tunnel vision. Saying that stimulating all the perceptible FOV
of humans is ineffective is like saying that surround sound doesn't work
for the ears. Granted, a vision application can take advantage of the
lack of visual acuity of the human eye to present lower resolution in
the periphery, but only as long as the eye is pointed straight ahead.
The higher resolution is needed whereever foviation occurs, so as the
eye moves, so must the higher resolution. I am a proponent for high
resolution insets about the line of sight with lower resolution outside
of the human "stationary field" (to use Sanders' terminology). The HMD
still needs the high resolution display over the entire "eye field"
(again borrowing from Sanders). Software or hardware can implement the
lower resolution presentation outside of the eye field if data bandwidth
is limited to the HMD.
In conclusion, I have found all commercial HMDs for which I have
experience to be adequate only for focused manipulation tasks. For
navigation, they are marginal, and for interraction with other persons
or moving objects, they are sorely inadequate.
That's my two cents, discounted for inflation.
Ian van der Linde wrote:
> > Bryn
> >
> > Yes, but let's not forget that the "data" part of a
> > display is only that part
> > which needs to be foveated. Peripheral vision is a
> > significant part of cueing
> > foveation. With properly coordinated head motion and
> > display update, a wider
> > field of view is desireable. In that respect, no
> > adequate HMD exists
> > commercially.
> >
>
> For manipulative tasks, it has often been demonstrated
> that limiting FOV causes degradation to dexterity and
> execution time. Although biological data suggests
> that we have a sensitive periphery, psychophysical
> testing indicates that reducing peripheral detail can
> be undertaken, such that the field of view is non
> uniform with respect to spatial resolution. This has
> been shown to have little effect on manipulative
> tasks,
> providing the resolution dropoff of the HMD is
> comparable to the fall in resolution of the retina
> with
> optical axis eccentricity. Beyond a certain
> eccentricity, I would say that the far peripheral
> field has an insignificant contribution to make to
> spatial awareness. It is better to employ this
> bandwidth for increasing temporal resolution for a HMD
> surely, rather than having a big FOV.
>
> Sure, large saccades may target features 700degs away,
> and limiting detail may obscure target stimuli, but
> the importance of this for most tasks is minimal.
>
> Ian van der Linde.
> Anglia Polytechnic University
> UK.
>
> ____________________________________________________________
> Do You Yahoo!?
> Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk
> or your free @yahoo.ie address at http://mail.yahoo.ie
--
Name : Bryn Wolfe
Title : Robotics Engineer
Dept : Texas Robotics & Automation Center (TRACLabs)
Company: Metrica, Inc
Voice : 281-461-7886
FAX : 281-461-9550
Web : http://www.traclabs.com
Email : mailto:[log in to unmask]
Smail : 1012 Hercules Drive
Houston, TX 77058-2722
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|