[opendtv] Why Europe should choose 720P for HDTV
- From: Craig Birkmaier <craig@xxxxxxxxx>
- To: OpenDTV Mail List <opendtv@xxxxxxxxxxxxx>
- Date: Sat, 11 Dec 2004 06:48:53 -0500
Donald Koelemanhas just learned that we moved OpenDTV to Freelists,
and has re-upped with the list. He asked me to post the following
provocative message while I am setting up his account.
As many are aware, the Europeans are now growing interested in
HDTV...again. As Alan Roberts has pointed out, there has been
considerable debate in Europe whether they should move forward with
1080i or 720P. The following piece delivers what I believe to be
convincing arguments for 720P from industry guru John Watkinson, from
whom I have learned nearly everything I know about digital television
and sampling theory (please no comments from the peanut gallery about
the depth of my knowledge on these subjects ;-)
So here's John's commentary, which I hope will promote so "new"
discussions on a very "old" subject.
Regards
Craig
P.S. and welcome back Donald!
>From Dave Sparks' The Prompt!!! A discussion piece by longtime progressive
scan advocate John Watkinson.
Cheers,
Donald
.....as does the 'Prompt!' 'HD in Europe' debate with this latest
contribution from a somewhat irked John Watkinson - author, lecturer and
digital guru. John writes:-
"Let us be quite clear that interlace is a primitive lossy compression
technique that allows a TV signal with an inadequate frame rate to give an
imperfect illusion that the picture rate is twice as high. All lossy
compression systems produce artefacts and in the case of interlace these
are quite serious. They include a substantial amount of residual frame rate
flicker as well as serious loss of resolution in the case of even very slow
subject motion. Unfortunately the loss of resolution is accompanied by
serious vertical aliasing.
"Thus you either have to categorise PAL as having 600ish lines of static
resolution and a picture rate of 25Hz, giving very poor motion capability,
or it has a picture rate of 50Hz and the resolution is halved.
"Static resolution is an outmoded and discredited metric of TV picture
quality. A bit like quoting how well the handbrake performs on an F1 car.
In all real TV programme material dynamic resolution gives a metric that
better represents perceived quality.
"I don't care for Mike Tooms' simplistic view of high definition.
Resolution is not the only parameter. What about the artefacts?
"When an interlaced picture is watched by a tracking eye, at a very low
vertical speed the lines in successive fields are superimposed on the
retina, leaving a subsampled image that is riddled with aliasing. Thus to
say that 720p doesn't have much more resolution than 625i is to be utterly
ill-informed. On real TV material, the dynamic resolution of 720p is about
three times that of 625i. In addition 720p doesn't flicker or alias. You
can tell an interlaced TV from a progressive TV several hundred feet away
because of the difference in flicker.
"Those who understand interlace know that it works better with a small
number of lines and a high field rate. This is why NTSC works so well. A
corollory of this is that interlaced high definition is an oxymoron.
"It's well known that you shouldn't concatenate compression schemes, so it
must be sub-optimal to put interlaced pictures into MPEG. You need a higher
bit rate with interlace, but you still get the interlace artefacts.
"Those who understand imaging will also know that the resolution of a
display is always less than the number of lines it contains because of
aperture effects. Thus a display with 1024 physical lines will be perfectly
matched to a 720p transmission by use of an interpolator. Interlace
complicates the design of interpolators.
"Consequently as far as I am concerned, the logical choice for future TV
broadcasts is 720p and this has been understood for several years now. The
continued existence of a debate on the subject simply illustrates the depth
of ignorance or the strength of vested interests, both of which I find
exceedingly tedious.
"If we accept that the HD viewer is supposed to sit closer, doesn't that
put more of the picture in peripheral vision, where the tolerance of
flicker is lower? Is this not an argument for increased frame rate? Surely
1080p is not the way to improve 720p. Far better to retain 720p but
increase the frame rate to 72 or 75Hz.
"Having seen such a system working I can vouch for the performance, freedom
from artefacts and general realism.
"Note that with a higher frame rate, temporal coding, which is where the
power of MPEG resides, becomes more efficient. Thus in an MPEG delivery
environment it is a myth that high frame rates cause difficulty."
I have a feeling John's comments may provoke some interesting reaction. And
you have all of the holiday season to compose your thoughts! (Next
'Prompt!' - 07 January 2005.)
----------------------------------------------------------------------
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:
