[opendtv] Re: Mediaflo Technology

  • From: Mark Schubin <tvmark@xxxxxxxxxxxxx>
  • To: opendtv@xxxxxxxxxxxxx
  • Date: Sat, 11 Nov 2006 20:53:01 -0500

Tom Barry wrote:
I guess another way of looking at it is the delivery resolution should be approaching (with diminishing returns) what we CAN see. But the display resolution should be safely into the zone of what we CAN'T see, letting our visual limitations furnish that free filtering for the tiny square pixels on fixed pixel displays. This again goes back to my liking for a 720p delivery but 1080p displays, at least when bandwidth is expensive.
That's not bad (1440p or more would be even better) for a situation in which the visual-acuity limitation is 720p. But visual-acuity limitations involve both resolution and contrast.

I'd love to attach a slide I created involving a contrast-resolution sine grating (contrast increasing towards the top, resolution increasing towards the right) with two curves superimposed on it, but I think Freelists won't allow attachments, so I'll describe it.

Near the bottom is a lopsided smile-shaped curve representing the contrast-sensitivity function (CSF) of a particular observer. The trough of the smile (maximum sensitivity, i.e., the ability to perceive minimum contrast) is at around three cycles per degree. The necessary contrast increases to the left (less resolution) and even more sharply to the right (greater resolution). Everything under that curve is invisible to an observer with that particular CSF .

Starting at the top left is a curve like the right side of a bell-shape. It's a modulation-transfer-function (MTF) curve. It ends up at the bottom right, but the exact shape of that curve is determined by such factors as the number of samples in the display. Anything above that curve simply isn't reproduced by the system.

The two curves intersect. Where the visual system needs the most contrast, the television system provides the least (i.e., where the CSF curve is highest, the MTF curve is lowest). There is a wedge at the left side of the grating, which is both visible (above the CSF curve) and reproduced (below the MTF curve). EVERYTHING we see from a particular television system must fit within that wedge. It's pretty scary. The first time I showed it to a top engineer in our field, he argued that it couldn't be, but eventually he accepted it.

We have no control over a viewer's CSF. The only way to increase the size of the wedge is to increase the MTF. More samples is one way to do that.


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: