Thanks, Craig.
A 60 Hz frame will contain the same blur as a 60Hz field. Assuming there is no
electronic shutter used, the motion in a typical frame of sports action is more
than enough to overwhelm any difference in the resolution of the image. It
works the other way, too. Early on, it was found that “jump and weave” in a
film projector can give back any resolution advantage that film might have over
video during projection (but that’s not really blur.)
In horse racing coverage (one of the original subjects of high frame rate video
replay) short shutters weren’t the answer, either. No one made a “slit scan”
camera for live TV in those days (which is what they really wanted.) High
frame rate cameras were what they used in film, and high frame rate video
cameras (and recorders) were developed to emulate that.
It’s not just ‘clueless’ broadcasters that struggle with 1/1001 frame rates.
The film folks have problems, too. Not to mention the audio and time code
headaches…
“9x16” cell phone video is more an artifact of form factor than it is an
artistic choice. People didn’t turn their old 35mm cameras sideways unless
they really wanted to. Just as giving people word processors doesn’t make them
talented writers, giving everyone a camera doesn’t make them photographers,
cinematographers, or directors.
I will leave your other disparaging remarks about broadcasters where they lie.
Regards,
Ken
From: opendtv-bounce@xxxxxxxxxxxxx <opendtv-bounce@xxxxxxxxxxxxx> On Behalf Of
Craig Birkmaier
Sent: Sunday, March 22, 2020 10:45
To: opendtv@xxxxxxxxxxxxx
Subject: [opendtv] Re: YouTube, Amazon Prime forgo streaming quality to relieve
European networks
On Mar 21, 2020, at 12:31 PM, Hunold, Ken <KRH@xxxxxxxxx<mailto:KRH@xxxxxxxxx>>
wrote:
I don’t often weigh in here, but I’ll break that rule just for now. Craig,
your opinions on interlace and broadcast television are WELL known, and
frequently cited.
Yes, I was indeed fortunate to be associated in the early ‘90s with some of the
finest minds in the advanced technology labs developing the technologies that
have changed the way we communicate. At the time, the global Broadcast TV
industry was trying to build a defensive wall to protect itself from new
competitors (cable and DBS), and the rapidly growing computer industry, which
had already transformed audio production and set its sights on digital video.
Interlace was viewed as one such barrier. If anything, I hope my work helped
educate the video industry about video compression:
- That interlace was nothing more than a crude compression technique that
proved useful in the era of analog television;
- That digital compression techniques offered far more powerful tools to reduce
entropy, in turn enabling vastly improved image quality at ALL levels of
resolution.
In fact, entropy coding improves in efficiency as both resolution and frame
rates increase. It is also fact that digital entropy coding of interlaced video
streams adds unnecessary complexity and reduces compression efficiency. Even as
digital compression has improved, at best we can “preserve” the loss of
vertical resolution and motion detail inherent with analog compression (i.e.
interlace).
The TV industry succeeded in crippling MPEG-2 by entrenching technologies for
which most patents had already expired, creating “new” intellectual property
that allowed their participation in the MPEG-LA royalty pool. And they
succeeded in protecting their investment in the 1125/50/60 HDTV formats that
sadly still exist today as 1080i.
“In the future, all displays will be progressive.” On this they were correct.
CRTs ruled the day for broadcasters, and flat-panel displays were in their
infancy.
Indeed! This legacy ALSO caused us to live with the CRT color gamut, delaying
the development of an extended color gamut by at least a decade. Not to mention
the development of an improved dynamic range (HDR).
It is well worth noting that Europe did not move to HDTV, choosing instead to
protect its investment in “601” digital gear which was both MPEG-2 MP@ML and
CRT friendly. By the time Europe started to adopt HDTV, flat-panel displays had
replaced the CRT as a consumer product.
“Future displays are not likely to be greater resolution than 1280x720.” On
this, they were wrong. As noted earlier, flat-panel displays were just being
developed. They were probably just 640x480 then. “HD” came later. Who would
have believed that a 4k display would be almost a commodity now?
We did! The 1992 SMPTE/EBU Task Force Report on Digital Imaging outlined the
digital imaging hierarchy that exists today. I would admit that we did not
predict 8K, but there is little practical application for this level of
resolution in consumer displays. Even theaters seem quite content with 4K with
HDR and ECG. 8K will find a place in image capture, as oversampling is critical
in the post-production process (think 35mm film).
“720p’60’ is better for sports.” My opinion is that this, too, is wrong. At
the time, “Super Slo-mo” replay devices were already doing 90 frames per second
(in 525i!) and that did make a difference. To me, there is no advantage of a
blurry 1280x720 frame over a blurry 1920x1080 field. For sports, you need more
frames, and camera/replay devices today have that covered nicely.
I have to take issue with some of this. 720P frames are not inherently blurry,
and 540 line fields do not carry as much vertical detail, although there IS
improved vertical detail when a 1080i “frame” is viewed, as long as motion is
minimal.
The reality is that we were still living with CCD imagers that were the major
limitation for both 720P and 1080i. The reality today is that most sports is
captured at 1080P then converted to 720P or 1080i for emission. I DO strongly
agree about higher frame rates, especially for slo-motion replay, which as you
note are in widespread use today.
You may finally get your wish to leave interlace behind as broadcasters produce
1080p or better. In the move to UHD, there probably isn’t a need to bring
1080i along with it.
The real question is whether TV broadcasters want to change anything. The ATSC
3.0 standard is little more than a novelty here in the U.S. Few broadcasters
want another major round of investment in an industry that is feasting on
retransmission consent payments from the competitors who deliver their content
to most homes in the U.S.
Now, if we could only get rid of 1/1001 frame rates. That’s another legacy
that still defies all efforts to remove it.
Yes, the broadcasters are still clueless about this. But they are more than
willing to use video captured by millions of smart phones, which do not support
the 1/1001 legacy.
Which reminds me about one other issue we addressed in the task force report:
ASPECT RATIOS
Who would have predicted that we would routinely be watching news clips
captured in the 9 x 16 aspect ratio?
;-)
Regards
Craig