[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: ATV / DTV Standards



I'd like to preface this by saying that I think we all understand that
nothing will change.  The standards are what they are and the market and
other forces will ultimately decide what happens and what survives.  In
these messages, for the fun of it and as a mental excercise, we are playing
"what should have been" and "what could have been".

> The only way out of this "chicken and egg" dilema is
>a bridge between where we are and where we're going.  So compatibility with
>NTSC is 100% vital now, because we'd never get to DTV without it.


Yes, of course, we all understand that NTSC/PAL are not going to dissapear
from one day to the next.  My line of reasoning was that compatibility with
the old standards shouldn't have a place in the new standard.  Compatibility
should be achieved via standards conversion.  If the standard defined, as I
put it before, a "digital film" analog, where one digital video
non-interlaced frame corresponded to one and only one film frame, then
conversion to NTSC and PAL would be very easy and of the highest quality.
Production and post would use these formats and the broadcasters could
invest in their converter of choice to be able to transmit DTV material in
the old standards.

I think that the whole thing could have been sold very well at the
production and post-production level simply on the basis of a greatly
simplified process flow, tape and disk savings and a less violent line of
communication with film-makers.  I've had meetings with film-makers with the
sole purpose being to explain why when going to video we have to "drop some
of their frames".  I don't blame them for not understanding.  All this stuff
would go away in my little world.

>It was realized that about one million new HDTV receivers per year was
about
>the most the consumer electronics industry could be expected to suppy, at
>least for the first few years.

Well, I don't know who came up with these figures.  The computer industry
produces and sells no less than 40 million PC systems a year.  Clearly the
production capacity exists.  Demand is another matter.  What I am saying is
that people are going out and paying $1000 to $3000 per system at the tune
of 40 million units a year.  The buying capacity and interest is there.  We
would have to offer the viewing public something that would motivate them
enough to go out there and want to spend $1000 or more on a TV set.   That
sort-of brings me to a fact that has nothing to do with the technical merits
of any standard:  Do you think people would go out in hordes to buy sets and
see Oprah (no offense intended, just the facts) in high-def?  Nope.  NTSC is
more than adequate for this type of programming and a good deal of what's on
TV today.
The only two segments of programming that could be used as bait are sports
and movies.  And, within those segments, only those viewers that are
hard-core fans and also appreciate the image quality gains.

Hence my thought that I don't see the problem as one of hardware production
capacity or standards compatibility but one of offering something that will
make people want to make the move.  Unless we ram it down their digestive
system they will not go out there and swallow the hook en masse.

On the other hand, nearly every PC and Mac out there is perfectly capable of
displaying at least something like from 480 to 720P very nicely.  Personal
computer people are technology nuts.  They jump on every new wave.  If you
offered them a DTV receiver card for $200 to $500 I'll bet you'd sell
millions of them a year.   The twist is that computer folks want more than
just sitting there and being fed.  They want to interact with it.  That's
where I see what the computer companies want to do as very valuable.  It
could push the adoption of DTV to a faster rate and provide a nearly instant
market.


>There are many conversions in digital
>video that are just as impractical to solve, regardless of the brilliance
of
>the engineers or the companies they work for.  Due to the temporal problems
>involved, I suspect that 29.97 -> 24 would be a pretty nasty one.


I was told that a box presented at ITS does a pretty good job of this type
of a process and de-interlacing.

Again, look at stuff like Oprah and tell me that there are issues.  Any
material that is critical will probably exist in film form and the owners
will probably choose to re-transfer and maybe even re-cut it to take
advantage of such things as the nicer aspect ratio.  I think that's the sort
of thing that you let the market deal with.
If you watch any amount of programming on one of the MPEG -based satellite
TV services you will see all kinds of compression and decompression
artifacts.  To us video folks that are used to the pristine 601 world the
stuff looks absolutely horrible and is hard to watch.  Guess what, to the
rest of the universe it looks great and they don't see the hits and misses
we pick up on in an instant.
So, I think a sense of proportion needs to be thrown in for good measure.


>The transmission aspect of DTV depends only on
>having signals that are amenable to MPEG-2 compression, not on the number
of
>lines used in a camera or telecine. So the guy who does 480p can get out
just
>as well as the guy who uses 1080i.  Whether or not you see a big difference
in
>your home depends on the TV set you buy.


True.  But I wasn't concerned with the abiltiy to transmit or not.  Being
that stations these days run on automation systems based on timecode, the
importance of a stable and consistent delivery format might be self-evident.
I don't even want to think about the nightmares of trying to make Louth deal
with 24fps, 25fps, 29.97fps, 30fps and 60fps timecode.  A common delivery
standard would at least simplify the problem to a single solution.

>But not everybody thinks film is good enough for their purposes. . .

I don't see anyone complaining about how little money Titanic, or Star Wars
or a whole host of other films are making.  Yes, a higher frame rate would
be great.  However, economics and reality will determine that most
programming will not need it or even justify it.  Take golf for example, is
anyone really going to get worked-up because of motion blurr of the ball at
24fps... probably not.

My thought was that the initial standard should have been something of a
more well defined product with a linear resolution scale.  Easy to
understand, easy to use, easy to sell and with no room for confusion or
conflicting choices (interlace or not?).  The standard would contain this
header I mentioned which would allow for future extensibility and options.
One such extension might be to take a 1/2 film res stream at nominally at
24fps and tell the receiver to interpret it as 1/4 res 48 fps.  So you'd be
able to capture, produce and provide programming at a higher frame rate in
the future.

Everyone wants to drive a Mercedes but only a few are willing to pay for it.

-Martin





---
Thanks to Queue Systems and Lipsner Smith for support in 1998.
No product marketing allowed on the main TIG.  Contact rob at alegria.com
995 subscribers in 39 countries on Tue Jul 14 12:41:22 PDT 1998 
subscribe/unsubscribe with that Subject: to telecine-request at alegria.com
complete information on the TIG website http://www.alegria.com/tig3/