[opendtv] Re: Is this Apple's iTV?

  • From: Craig Birkmaier <craig@xxxxxxxxx>
  • To: opendtv@xxxxxxxxxxxxx
  • Date: Sat, 17 Mar 2012 10:52:43 -0400

At 3:00 PM -0500 3/16/12, Manfredi, Albert E wrote:
Cliff Benham wrote:

 OK, what percentage of 1080p or higher resolution programming from
 all sources and transmission schemes is actually available to
 consumers to watch at home in the world today? What I mean is how
 much 1080p is actually getting to the inputs on my HDTV?

Cliff, I had a couple of reactions to this better than 1080p resolution of the new iPad.

One reaction was the same as yours. The other reaction was, watch Craig gush all over this, in spite of the fact that he poo-poohed HDTV from day 1.

Not a fair representation Bert. I poo-poohed the ATSC standard, the continued use of Interlace and the notion that everyone would be required to buy a receiver that most would not use. And I poo-poohed the notion that HDTV would protect broadcasting from the relentless assault of Moore's Law.

The fact that we are now talking about true 1080P displays and the streaming of 1080P content is more than adequate vindication...

The reason I didn't comment was that here is Craig's likely response: "It doesn't matter whether the source material is 1080p or better. You still gain image quality by oversampling the image to the higher resolution."

There are two issues at play here Bert, and they are both related to sampling theory.

Video is Nyquist limited - it must be filtered properly to display smooth motion without aliasing. As a result, the video resolution that can be delivered to any progressive display will necessarily be half or less of what the display can deliver in terms of Non-Nyquist filtered imagery. AND, since the digital delivery system is based on entropy coding techniques (digital video compression), the actual resolution of the source may be much lower than the theoretical, especially when the channel bandwidth is constrained.

Thus, as I have stated many times before, you may get a higher quality image on the display by upscaling a lower resolution version of content where the samples are not compromised, than a higher resolution version where the samples have been trashed by too much compression.

Bottom line, the resolution of the display is only one factor in delivered image quality, but a higher resolution display can do a very good job of presenting ANY source by upscaling.

Now for the more important issue.

Most of what we view in our browsers and APPs is NOT Nyquist limited. The samples are addressed directly - you can place black pixels next to white pixels...

I have been using a retina display for more than a year - on my iPhone. The ability to render very fine details with high contrast has been a breakthrough for the content I view the most on this screen.

Consider this. The Desktop Publishing revolution was driven by the ability to use a relatively low resolution representation of a page (~ 72 DPI) to create and proof a printed page. You could even print out this representation on the best computer printers of the day - a dot matrix printer. Then the Laserwriter changed the landscape; you could print out a much higher resolution version of the page (300 DPI). The quality was nearly as good as professional printing, with crisp text and graphics WITHOUT visible aliasing. OR you could output to a digital pre-press system at resolutions up to 2500 DPI.

What the retina display does is deliver near laser printer quality to the screen, with very high contrast ratios. Text looks as good as a printed book; photos look like high quality prints. And the color gamut is wider than that used for HDTV formats; this means better color saturation and a wider range of colors.

I went to Best Buy yesterday to look at the new iPAD. It does for a tablet exactly what the retina display did for the iPhone. This is one of the best displays I have ever seen. And the extra pixels (relative to HDTV) play an important role. The 4/3 or 3/4 screen aspect ratio (remember you can rotate the tablet to view in portrait or landscape mode) is friendly to all kinds of content, especially web pages, books and photos. And the high resolution makes watching letterboxed HD very enjoyable (or you can zoom to fill the screen).

I suppose that tablets are held up to viewing distances more or less like those of HDTV sets, in terms of picture height viewing distance. So, there's probably some rationale for HDTV quality images, at best. IMO, much better than that is wasted. But if Apple does it, I suppose we all genuflect and Believe.

In the '90s we talked about viewing distance for different applications. Computers were up close and personal - 15 to 30 inch viewing distance. TVs were for couch potatoes - generally 7-9 feet minimum viewing distance. It was widely believed that people would not watch TV on their computers sitting in an office chair.

The table as a "second screen" changes the dynamics. There was a very important reason that when Apple introduced the iPAD, Steve was sitting in an easy chair. The table can be used as a lean back device for media consumption, game playing etc. One of the most popular viewing venues for the iPAD is in bed.

The fact that the iPAD screen is a good HDTV screen at the nominal viewing distance is important; but the fact that it is ALSO able to deliver stunning imagery for other applications is equally important, if not more so...

Think about the impact of a doctor making rounds and pulling up high resolution medical images on that 264 DPI screen...


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: