[opendtv] Re: Popular screen aspect ratios

  • From: Craig Birkmaier <craig@xxxxxxxxx>
  • To: opendtv@xxxxxxxxxxxxx
  • Date: Fri, 7 Jan 2011 07:24:16 -0500

At 5:05 PM -0600 1/6/11, Manfredi, Albert E wrote:
That's pretty funny. You have applied the proper aspect ratio accommodation, so you will never need to?
Think of it like a railroad freight car Bert. A standardized container with which you can carry all kinds of content.
I am not saying this was the best outcome, just that this is the reality. The best outcome would have been to implement the compression standards as intended. That is, you compress the source as is, no matter what the resolution and aspect ratio, and signal what you have compressed; then you let the decoder/display use the metadata to optimize the presentation on that display.
I think the problem is, you don't understand the difference between the input signal to the MPEG-2 decoder, and the output signal from box to display. And I think that has been your misunderstanding from the start, back in the 1990s.
The MPEG-2 metadata, that you thought was the whole solution, only applies before the decoder does its thing. The problem has never been there.
The decoder needs to know how correctly fill the image frame in the display, to avoid distortion. This has to occur BEFORE the end user plays with his zoom or other display features. The baseband analog signal to the display cannot convey this information. It doesn't have MPEG metadata. THAT's where the problem was and is.
No Bert, the decoder does not need to know ANYTHING about the display. It simply needs to know how to reconstruct the source. It is the job of the display processor to take the metadata about the resolution, decoded pixel aspect ratio (in the case of an anamorphic squeeze), and intended display aspect ratio, and scale the decoded source for the local display. This is exactly the way it works with PCs.
The reason that "some" decoders also must deal with display accommodation is quite simple; some displays cannot do this, like legacy analog TVs. And to make things worse, the dumb TV crowd wanted to create ONE FORMAT for everything...AGAIN. So the net result is that we have one container into which content producers are putting a wide range of formats, instead of simply relying on intelligence to deliver anything and let the display deal with it.
And if there were 21:9 decoders suddenly on the market, it wouldn't work! Not over analog interfaces, not even over HDMI.
This is absurd. It already works all the time for computers. And it can EASILY be signalled over HDMI; the only limitation is the capability of the display processor to work with larger rasters, which the 21:9 Phillips set clearly can do.
I do not understand why you resist such a simple idea, that is the basis for all new entropy encoding standards. These standards could care less about formats.
So, the comment "we don't need standards anymore" was nonsense in the mid 1990s for sure, and continues to be wrong today. It may have been otherwise with HDMI, but it appears from 

http://en.wikipedia.org/wiki/Extended_display_identification_data
that HDMI only supports 4:3 or 16:9. In many cases, pixels are not square. Scroll down to the table "CEA EDID Timing Extension Version 3 data format."
This is largely irrelevant. It is a standard for connecting displays to computers and other devices that may be based on specific video and audio formats.
As for auto identification and setup you may note that there is a field for Manufacturers reserved timing for a display. So a display of any resolution and aspect ratio can be identified.
The table of video formats is related to the need to support a range of external video sources that might be connected directly to a display, such as a projector. This is analogous to fields 35 and 36 Established timing I and II. The standard identifies a bunch of known stuff, but also supports variables for other formats.
The ability of the computer - connected to the display - to decode any random raster of video is completely unrelated to this standard. Once the computer is properly connected to the display it can do anything to fill that screen the way it wants to. This includes decoding any video format and accommodating it to the display. All the PC needs to know is what is already identified via the VESA EDID DATA:
1. The actual raster size of the display; (note it is probably assumed that all computer displays are square pixel). 
2. The aspect ratio of the display.
An interesting note here, the standard even identifies that actual screen size, so it is possible to know exactly how large the image will be and adjust the scaling if necessary. 


Give it up, Craig. Simplistic notions are almost always wrong.


This is not a simplistic notion. It's the way things work!
Please give up trying to make this more complicated than necessary. There is no benefit from having a few fixed formats. 

Regards
Craig


----------------------------------------------------------------------
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:

  1. Home
  2. opendtv
  3. Archive
  4. 01-2011