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

DTV News and other bits



I am cross-posting this as I think it is of interest to many here who may not
read the other lists. If anyone would like me to stop doing so, please let me
know!

Tom

DTV News:

- Oops Department:  Today's Consumer Electronics newsletter reports
problems with first-generation DTV receivers.  The most serious appears
to be the differential phosphor luminance decay (DPLD) problem that so
many said wouldn't happen.
     In stores running 16:9 video (from Sencore servers) almost
constantly on 4:3 Mitsubishi projectors, stripes were burned in. 
According to Martin Zanfino, Mitsubishi's product-development manager,
"Our dealers ran them for a very long time during the day, so that
accelerated anything that would happen to a consumer, and retailers
aren't mixing the [16:9] programming with anything else.  We don't
believe that a customer that watches 4:3 now and then switches to an
HDTV broadcast or DVD with a 16:9 aspect ratio will see anything wrong."
     It's not just Mitsubishi and DPLD.  Here's one retailer on a
problem with Sony DTV sets going black and/or silent:  "The way in which
some of the stations around the country broadcast the high-definition
signal poses a problem for the Sony sets, but it can be repaired
easily."  And there's more:

- The audio/video-sync problem reported earlier this month at the
Advanced Television System Committee's (ATSC) T3 technical committee is
apparently both real and widespread.  That was ONE of the tidbits that
emerged from last week's ITS Technology Retreat in Palm Springs.
       The A/V sync problem was acknowledged by at least three networks,
and it was said to be as bad (worst case) as two-to-three SECONDS and to
sometimes vary both with channel changing and when some sets are turned
off and back on.  Given that no receiver manufacturers were directly
represented at the event, it's not too much of a surprise that they were
blamed and the transmission system was exonerated.  Even so, those
blaming the receivers acknowledged that receiver manufacturers had told
them that they had tested lip-sync with the Sarnoff lip-sync test
streams; they were also hard pressed to explain why the problem was
manifested on so many different receivers if it was a receiver problem.

- Other tidbits from the ITS event:
     - There appears to be a strong desire for an ATSC-emission
simulator to be used by post houses to monitor quality.
     - There appears to be a great deal of interest in SMPTE's metadata
dictionary work.  Bill Miller (SMPTE engineering vp) gave a mini-session
on new, low-cost ways of participating in SMPTE's standards work, which
seemed to be appreciated.  Contact him by e-mail at
William.C.Miller at abc.com.
     - There appears to be a desire to be able to do multi-pass
encoding, as in DVD, for maximum quality, somehow inserting those
streams into the ATSC multiplexer post encoding.  These comments first
emerged after Richard Mizer's presentation on the difference between
program streams in DVD and the ATSC's transport stream.  I don't see
this happening for a  l  o  n  g  time.
     - A number of eyes and ears were opened by some comments, such as
that of experienced HD shooter/engineer C. R. Caillouet that one cannot
do critical HD focus on a 1.5-inch viewfinder.  John Sprung's
demonstration of experiments in dual-aspect-ratio accommodation at
Paramount also seemed to elicit some "ooh"s and "oh"s.  His split-screen
demo of expansion from anamorphic 810 lines, however, which seemed to
indicate an (appropriate) inability to tell one side from the other at
the ITS Forum last summer and the SMPTE convention in October, was
easily discerned by almost everyone in the Technology Retreat demo. 
Better projector?  Better audience?  More experience?
     - Other tidbits (my opinions) from the demos (MUCH appreciation to
Bruce Jacobs -- and the manufacturers -- for arranging everything):
          - There is still something very attractive about a plasma
display.  720p material on the Pioneer 1280 x 768 looked great, even
though it clearly didn't have the contrast of some of the other
displays.  Sony's 852 x 480 display also looked pretty good -- better
(to me) than the Philips display based on what was said to be the same
Fujitsu panel.
          - There are interesting problems with internal scaling engines
in displays:
               - One demo involved a split-screen between 720p original
material and the same material downconverted to 480p and upconverted
back.  On most of the displays -- even the 480-line panels -- the
original, as one would expect, looked better than the round trip.  But a
JVC plasma panel had trouble, apparently, with the high frequencies of
720p.  On THAT display, and that display ONLY, the down-and-up looked
better than the original.
               - There were also some noticeable problems with the
Hughes/JVC D-ILA projector's internal scaling.  This one had supposedly
been optimized on behalf of the Department of Defense for 720p, and it
certainly looked better on that format than what I had seen previously
on a different D-ILA.  The person who brought it attributed this to
another political issue associated with the 1080i camp (see later in the
Panasonic story).  Still, there were some artifacts visible on some
critical program material.  We started playing with the various
parameters, and it's possible that, with enough time, we might
eventually have come up with some size and pixel array that would have
worked better, so take this one with a grain of salt.
          - The CRT-based projectors and displays didn't have scaling
problems, but they DID have scanning problems.  Someone commented that
he thought that not a single one of the 12 displays out front (or,
probably, any of those in the back) seemed able to show pictures on
EVERY format demonstrated.  Some seemed unable to handle 720p; others
gagged on 480i.  Some seemed able to handle 1035i from HDCAM or the 
HD D-5 but not, for some strange reason, from the HDD-1000.
          - On ordinary, sitcom-ish program material, all of the down-
and upconverters seemed pretty close to one another in performance,
though one might choose one or another on the basis of subjectively
"punchier" pictures.  On more critical material, however, edge and other
high-frequency effects were more noticeable.  The same was true of test
signals.  What one might choose on the sitcom material might not be what
one would choose on the critical material.
          - One sequence was a tree stump decorated as an elf's house. 
It had originally been shot 1035i and was downconverted to 480i, which
(not having the original to compare it with) looked fine.  The
upconversion, right next to the "original" in 480i, however, looked
soft.  I didn't find that so unusual, believing in Charlie Pantuso's
dictum that everything looks best displayed the way it was intended. 
C.R., however, wanted to know why it looked so soft.  We adjusted many
monitor controls and discussed such things as the self-sharpening effect
of visible scanning lines and the smoothing effect of interpolation,
but, as of The Morning After event on Sunday, we still weren't
satisfied.  Good!  That's how we learn.
          - Everyone loved Bruce Jacobs' test material from the PBS
Resolution Rendezvous.  Bruce Penney of Tektronix liked it so much that
he recorded some of it for use in a SMPTE working group (someone
commented that the ITS Demo Day was like a mini-SMPTE conference, given
who was there and what was going on).
          - My sense was that everyone was happy with the look of Fox's
24p material on sitcoms and dramas.  Most people, however, were NOT
happy with the look of 24p or 30p sports.  I didn't mind it, but I'm not much
of a sports fan.
          - As usual, viewing angle played a big role in artifact
visibility.  What I had thought were glaring problems of the use of
non-HD lenses, when I had seen the material on the giant screen at Sony
Pictures, were barely visible on Demo Day unless viewed closely. 
Similarly, the problems caused by emission coding of high frequencies
needed close viewing.
          - Someone remarked that when he saw Panasonic people trying to
make Sony's equipment look as good as it could during the setup (is THAT
what they were doing?), he was proud of our industry.  The cooperation
WAS pretty terrific, and the setup was astonishing, not even counting
the fact that we had access to the room beginning at about 5:45 pm the
night before, after the last wedding guests trickled out.  Get hold of
the block diagram from Monica (mmathis at erols.com) or Bruce
(bjacobs at ktca.org) if you want to be blown away.  Yes, a major
multiformat facility, complete with HD-SDI routing and multiple format
converters, was put together in a few hours.  Yay, team!
     - Whatever one can say about their products, some of the members of
the format-converters panel seem to need a little tutorial on their
processes.  Asked technical questions about filtering and colorimetry,
they (not all) fumbled quite a bit.  One thing that seemed unanimous on the
eight-manufacturer panel (including Sony), however, was that format
conversion from progressive is always easier.  How much easier?  One
manufacturer said 80% of the current drawn by his machine is devoted to
nothing but deinterlacing.  In the audience, by the way, was a
representative of a format-converter-manufacturer wannabe, TeraNex,
which plans to have Lockheed-technology-based, parallel-processing,
motion-adaptive converters.  They'll have a suite at NAB.  I asked if
they could do temporal interpolation from 24p to 60.  The answer was
yes, but some potential customers they spoke to weren't sure they wanted to
lose the 3-2 pulldown ("film look"). Personally, it can't go away fast
enough for me.
     - Speaking of 24p, there was a great deal presented on that and on
Sony's segmented-frame (sF) plan.  It came out that Fox has been using
the same thing (what they call field-segmentation) for the 15 hours a
week of progressive DTV material they've been airing.
     - Sony announced plans to have 720p product next year (and Larry
Thorpe actually appeared to speak negatively of interlace!).  Five
manufacturers described different ways they will offer HD telecines. 
Philips will have a virtual telecine (many disk drives) and is reviving
D-6, which, in the 1080/24p mode, will have 12 uncompressed audio
channels.  The price is down to $160,000, and there will be a cap on
maintenance costs, after which it's Philips' problem.  There will be a
range of D-6 products coming, for anything from HDTV to data, the latter
falling into Philips' strategy that recording film data files is even
better than 24p.  The rates and capacities exceed even Ampex DST.
     - There were some lively audio discussions, too, including stuff
about how to measure loudness and how to monitor (and otherwise deal
with) surround sound.  One tidbit:  Some decoders don't offer subwoofer
outputs, and some movie mixes (including one involving dinosaurs) put
ALL the thumping in there, leaving those with such decoders to wonder at
how well creatures that large could tiptoe.
     - Dolby's Steve Lyman said there WERE DTV receivers with
dual-stream decoders (going so far as to interrupt one of his company
colleagues on he subject), but he couldn't say who they were.
     - The Department of Defense provided some color glossy surveillance
photos shot from a moving blimp with a commercial 720p camera.  They
were pretty amazing!
      - Television Broadcast, Videography, and www.digitaltelevision.com
sponsored Craig Birkmaier's The Morning After the morning after.  I
could only attend a little, but, from what I heard, it went well.  
     All in all, it seems to have been a pretty terrific Technology
Retreat.

- Just before the Retreat, Panasonic held its pre-NAB press conference. 
I got a preview the night before.  Here are some DTV-related tidbits:
     - Like everyone else, Panasonic will introduce 24p product, in this
case an HD D-5 that'll do just about any format known to have ever
flashed through anyone's brain, including 25p, 25i, and 720p/60.  The
only thing it WON'T do is have segmented frame input or output (but Sony
announced that a number of manufacturers will be offering converters). 
It also won't connect to anything else Panasonic makes when it's in 24p
mode.
     - Other DTV product highlights:
          - HD switcher and DVE
          - HD cart machine
          - DVCPRO HD (formerly DVCPRO 100)
          - 16:9 laptop editor for DVCPRO 50
          - a DTV board for Compaq
     - What's DVCPRO HD?  Good question, and I got different answers
from different Panasonic people at different times.  It's meant to
compete with HDCAM; it will use DVCPRO cassettes, which will have half
the capacity of DVCPRO 50 or a quarter of DVCPRO; it will record 1080i
initially; and it will use a new chipset (not ganging DV chips
anymore).  On that there was general agreement.  As for the rest, it may
or may not be compatible with JVC's Digital-S 100 (on a signal basis --
obviously, the tapes won't be compatible), and it may record 1280-pixel
luminance and 960-pixel chrominance, or maybe 1280/640, or maybe
1440/640.
     - Panasonic will also have a 1920 x 1080 plasma panel in a suite. 
Likewise, for the second year in a row, a 720p camera hidden in the
suite.  Why not put the camera in the booth on the floor?  "Commercial
reasons."  When I expressed some doubt, that changed to
"political-commercial reasons."  Those political reasons seem to be
holding up 720p equipment, which is why Sony's announcement was so
encouraging (although Sony has been known to slip on timetables in the
past -- remember the SMPTE '97 HD streamer?).

- The New York Times reported Wednesday on an agreement between IBM/NEC
and Hitachi/Pioneer/Sony on digital "watermarking."  The so-called
Galaxy Group watermark is said to be able to survive compression,
decompression, and even analog recording (though a Group-member's
representative on the ITS Technology Retreat format conversion panel
seemed a little skeptical about its surviving format conversion).  It's
supposed to be able to run an average of eight hours a day for 22 years
before yielding a false positive.
     There remains a second watermarking proposal from Digimarc,
Marcrovision, and Philips, and a decision on which to use is not
expected until summer.  This is in addition to the five different
proposals being evaluated for copy protection.

- Monday's New York Times "Business Day" section has a front-page story
on digital cinema.  "Within two years, movie theaters are expected to
begin installing the first generation of digital projectors."



---
Thanks to Bill Abbott for support in 1999.
No advertising/marketing allowed on the main TIG.  Contact rob at alegria.com
anonymous messaging now at http://www.alegria.com/HyperNews/get/ubique.html
1032 subscribers in 41 countries on Tue Feb 23 11:43:48 CST 1999 
subscribe/unsubscribe with that Subject: to telecine-request at alegria.com
complete information on the TIG website http://www.alegria.com/tig3/