[opendtv] Re: 1080p @ 60 is Next?

  • From: "Manfredi, Albert E" <albert.e.manfredi@xxxxxxxxxx>
  • To: <opendtv@xxxxxxxxxxxxx>
  • Date: Wed, 16 May 2007 17:56:13 -0400

John Shutt wrote:

> Based on what?  An arbitrary screen size viewed at an arbitrary
> distance?

I keep seeing claims that 720p is as good as anyone should need, but
these perceptual things are not to be taken too religiously, IMO. Real
world tests sometimes give unexpected results.

Just doing the numbers, with a simplistic model, 720p does seem
adequate, for 3 picture height seating.

You see claims that human vision resolves down 2 arcmin or 1.6 arcmin,
depending on the source. That's 0.0267 degrees, at most, for "normal"
eyes.

http://en.wikipedia.org/wiki/Human_eye

just over half way down the page.

Okay, so take a 50" screen, dimensions 24.6" H X 43.5" W.

Everyone seems to say 3 picture heights distance for HDTV. So that 50"
screen sits 73.8" or 6.15' distant. That sounds close enough to me, for
a 50" screen.

Assume you can perceive 0.0267 degrees, so 0.0343" separation between
pixels is the closest one can perceive on this 50" screen.

If the horizontal vertical dimension is 24.6", the distance between
pixels is:

0.0228" if 1080 pixels, and
0.0342" if 720 pixels.

Since 0.0343" was the limit we figured could be perceived, that says
that the 720p is just right for human visual acuity, and 1080p is more
than you need, for 3 picture height distance.

So if you really believe that visual acuity can be determined this way,
and there's nothing more to it, then 720p is it. But if real-world tests
show something different, then there are other mechanisms at work that
this simple model doesn't consider.

Perhaps the fact that the image can be scanned multiple times, for
example, rather than just fixed upon, makes the difference. If a part of
the image is sampled multiple times, it is not hard to believe that the
eye-brain may be able to detect edges that are less than 0.0343"
ambiguity.

So, IMO, it's pointless to get hung up one one answer and ignore tests
that might dispute that result.

Bert
 
 
----------------------------------------------------------------------
You can UNSUBSCRIBE from the OpenDTV list in two ways:

- Using the UNSUBSCRIBE command in your user configuration settings at 
FreeLists.org 

- By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word 
unsubscribe in the subject line.

Other related posts: