HDTV production equipment ???
d...@iceinc.com
I'm researching hdtv to produce digital film. I hear rumours of
progressive scan 24fps hdtv cameras. anybody know?
In my thinking you don't need progressive scan if you can let the
electronic shutter do the work for you. I've looked at specs on the
web for a couple of hdtv cameras. I haven't seen any with shutter
speeds less than 60 fps.
I guess I have to wait a few years or till Lucas does something about
it.
manf...@arl.bna.boeing.com
In article <331506...@interlog.com>,
Just out of curiosity, are you trying to shoot the film digitally and
then produce from it a standard 24 fps 35mm film? Otherwise, why would
you care whether you're shooting a digital film at 30 or 60 fps rather
than 24 fps? Isn't 24 fps in the standard _only_ to be backward
compatible with existing film material?
I also don't understand the comment "you don't need progressive scan if
you can let the electronic shutter do the work for you." How does the
shutter affect what kind of scan you are recording with?
Bert
manf...@arl.bna.boeing.com
-------------------==== Posted via Deja News ====-----------------------
http://www.dejanews.com/ Search, Read, Post to Usenet
David Richards
In article <8570572...@dejanews.com> manf...@arl.bna.boeing.com writes:
>In article <331506...@interlog.com>, d...@iceinc.com wrote:
>> I'm researching hdtv to produce digital film. I hear rumours of
>> progressive scan 24fps hdtv cameras. anybody know?
Yes, there is a high-def proscan camera available from Polaroid. There are
also some digital still cameras becoming available that use the 640x480
format, and output in proscan.
>> In my thinking you don't need progressive scan if you can let the
>> electronic shutter do the work for you. I've looked at specs on the
>> web for a couple of hdtv cameras. I haven't seen any with shutter
>> speeds less than 60 fps.
>
>I also don't understand the comment "you don't need progressive scan if
>you can let the electronic shutter do the work for you." How does the
>shutter affect what kind of scan you are recording with?
I think the other poster was assuming that with electronic shuttering, the
entire image is captured at the same instant, thus the odd and even lines
would not have a time interval between them. That is not the case. Even
with electronic shuttering, the odd and even scans are still captured
1/60s apart. An interesting question is this: If an image is captured
progressive, at 30Hz, how noticeable are the artifacts when the output
is viewed on an interlace monitor? (obviously there are no artifacts if
the source is captured at 60Hz, and the extra lines are simply discarded).
Dave
--
I do not appreciate unsolicited commercial email. If you send it to me,
I WILL complain to your ISP. Then I will hunt you down and kill you.
Louis A. Carliner
dave...@netcom.com (David Richards) wrote:
>Dave
My guess is that that artifacts will be absent. Let's see to it that
progressive CAPTURE be a requirement for HDTV, with interlaced
transmission fine for distribution. Decent quality de-interlacing
processing is horribly expensive, as with the case of the Faroudja
products, which continue to set standards. Also, progressive capture
will compress much better without objectionable introduction of
artificats or loss of video information.
Louis A. Carliner
Your source for video calibration expertise for
Washington, D.C., Maryland, Virginia and West Virgina
as well as Southern Delaware UNTIL LATE 1997
in transistion to West Central Florida
by an ISF/Joe Kane trained specialist equipped with both
the Philips color analyser and ISF optical comparator.
WILL TRAVEL ELSEWHERE! JUST PAY MY WAY!!
Phone: (301) 340-6120
email: lcar...@idsonline.com
Louis A. Carliner
dave...@netcom.com (David Richards) wrote:
>I hear rumours of
>progressive scan 24fps hdtv cameras. anybody know?
>Yes, there is a high-def proscan camera available from Polaroid. There are
>also some digital still cameras becoming available that use the 640x480
>format, and output in proscan.
> In my thinking you don't need progressive scan if you can let the
> electronic shutter do the work for you. I've looked at specs on the
> web for a couple of hdtv cameras. I haven't seen any with shutter
> speeds less than 60 fps.
>
>I think the other poster was assuming that with electronic shuttering, the
artal...@aol.com
> Let's see to it that progressive CAPTURE be a requirement for HDTV, with interlaced transmission fine for distribution.
So you suggest that we not use the current HDTV equipment that is now in second or third generation (analog and interlace) but instead wait until someone builds the progressive equipment in a few years?
Who do you think should set this requirement?
Production with modified HDTV equipment and conversion to digital before broadcast seems to offer a much more rapid path to providing new HDTV programing for the public. The progressive HDTV technology will compete on a cost/performance basis and survive or not in the free marketplace. Can you offer a reason why it should be otherwise?
Art
::{)
Charles A. Poynton
In article <daverichE...@netcom.com>, dave...@netcom.com (David
Richards) asks:
> If an image is captured progressive, at 30Hz, how noticeable are the
> artifacts when the output is viewed on an interlace monitor?
In article <5f6ofg$k...@nntp.idsonline.com>, lcar...@ids2.idsonline.com
(Louis A. Carliner) wrote:
> My guess is that that artifacts will be absent.
I commend Louis A. Carliner for pointing out that he is guessing. Many
Usenet contributors don't bother telling us when they're guessing!
In fact, there will be artifacts, upon display, on any scene element in
motion (relative to the frame) that is eye-tracked by the viewer. The
magnitude of the artifacts will be proportional to the amount of motion
(expressed in degrees per second, say). This is a display artifact, not
represented in the captured images.
About two years ago there was a thread on this topic, and I invited
readers to analyse why text - the message "EAT AT JOE's," say - scrolling
horizontally across an LED display appears tilted, despite the LEDs being
arranged in orthogonal rows and columns. I delivered a paper on this topic
a year ago at the SMPTE conference in Seattle: "Motion portrayal, eye
tracking, and emerging display technology." This paper was published in
the Proceedings of the conference; the Proceedings are available through
SMPTE. This paper is not available online, though public pressure might
force me to put an HTML version up on the web.
Cheers,
C.
-----
Charles A. Poynton
<mailto:Poy...@Poynton.com> [Mac Eudora/MIME/BinHex/uu]
<http://www.inforamp.net/~poynton/>
-----
kil...@aa.net
> >>In article <331506...@interlog.com>, d...@iceinc.com wrote:
> >>> I'm researching hdtv to produce digital film. I hear rumours of
> >>> progressive scan 24fps hdtv cameras. anybody know?
> >>> In my thinking you don't need progressive scan if you can let the
> >>> electronic shutter do the work for you. I've looked at specs on the
> >>> web for a couple of hdtv cameras. I haven't seen any with shutter
> >>> speeds less than 60 fps.
<many snips>
> > Even
> >with electronic shuttering, the odd and even scans are still captured
> >1/60s apart. An interesting question is this: If an image is captured
> >progressive, at 30Hz, how noticeable are the artifacts when the output
> >is viewed on an interlace monitor? (obviously there are no artifacts if
> >the source is captured at 60Hz, and the extra lines are simply discarded).
The CCD does grab a full frame each 60th (NTSC) of a second, but it sums line
pairs to create and read out synthetic odd or even lines to support interlace. It
is also a way of filtering the resolution from 480 lines per frame to as low as 240
lines per frame to make the interlace display scan watchable (but so much for
the myth of the Kell factor and 480 scan lines resolution).
If I follow the thread here, the question implied first was why not do film rate
video capture, and then why not do 30 frames per second using all 480 lines.
Shouldn't that be better than film? (faster frame rate = better motion, and the
same 480 real lines per frame available)
The problem with that is film blur vs. the relatively instantaneous video image.
Thirty frames/sec of video capture looks quite jerky with fast motion, camera
pans, etc., while with film it would look fairly good because the motion is
integrated over time, or blurred, which makes it look smoother. If you can
imitate film response electronically, you should get similar results to film (and
you should file a patent).
> My guess is that that artifacts will be absent. Let's see to it that
> progressive CAPTURE be a requirement for HDTV, with interlaced
> transmission fine for distribution. Decent quality de-interlacing
> processing is horribly expensive, as with the case of the Faroudja
> products, which continue to set standards. Also, progressive capture
> will compress much better without objectionable introduction of
> artificats or loss of video information.
> Louis A. Carliner
I second that notion.
>>>>
Kilroy Hughes
mailto:kil...@aa.net,
http://www.aa.net/~kilroy
Kilroy...@Compuserve.com
Future Media Systems, Seattle
kil...@aa.net
> artal...@aol.com writes:
> In article <5f6oan$k...@nntp.idsonline.com>, lcar...@ids2.idsonline.com (Louis A.
>
> > Let's see to it that progressive CAPTURE be a requirement for HDTV, with interlaced
transmission fine for distribution.
>generation (analog and interlace) but instead wait until someone builds the progressive
>equipment in a few years?
A fraction of a percent of existing production facilities are "HDTV" now. It isn't too late to do
it right.
> Who do you think should set this requirement?
The ATSC should be active in recommending, but the way it works is: consumers buy
receivers and displays that are proscan or not, and anyone who wants to broadcast to them
gives them what they want, if you don't you're out of business.
The process is bound to be quite democratic and hellatious in the US. I don't think a
"requirement" will have much to do with it. You won't get a big chunk of the population to
follow the MUSE into the ocean just because she's playing the official tune. (No disrespect
intended to the pied piper of Hamlin, mythological Greek characters, or Japanese early adopters
by that metaphor.)
> Production with modified HDTV equipment and conversion to digital before broadcast
seems to offer a much more rapid path to providing new HDTV programing for the public. The
progressive HDTV technology will compete on a cost/performance basis and survive or not in
the free marketplace. Can you offer a reason why it should be otherwise?
>
All the shows and movies (except sports) that are prime candidates for HDTV are shot on film.
Even prime time TV. Just broadcast that at 24 frames proscan. Put a Faroudja in the signal
chain to step the other 60 field video sources up to 60 frame before encoding and we're in
business.
36 frame per second and 60 frame per second (720 line) cameras and VTR's or disc
recorders should be deployed for football and basketball games sooner the better. When the
high res consumer display technology is cost effective in a few years and people decide to
start watching TV from one screen height distance, then we'll need a thousand scan lines.
Craig Birkmaier
> I think the other poster was assuming that with electronic shuttering, the
> entire image is captured at the same instant, thus the odd and even lines
> would not have a time interval between them. That is not the case. Even
> with electronic shuttering, the odd and even scans are still captured
> 1/60s apart. An interesting question is this: If an image is captured
> progressive, at 30Hz, how noticeable are the artifacts when the output
> is viewed on an interlace monitor? (obviously there are no artifacts if
> the source is captured at 60Hz, and the extra lines are simply discarded).
>
> Dave
There are two factors that affect the delivered quality in the scenario
described by Dave above.
First are the temporal rate issues. Capturing 30 temporal samples per
second clearly does not provide the same motion reproduction quality as 60
temproal samples per second. This relates to the truely inspired
compromise that helped make interlaced NTSC work so well. Thus if there
are objects moving at higher temporal frequencies, we may see strobing
artifacts in material acquired at 30 Hz--especially with side to side
motion and pans at certain beat frequencies.
Second, are the spatial resolution issues, especially relating to the
filtering used during image acquisition. If the images are filtered
vertically, as they are today for NTSC distribution--typically at 50-60%
of the potential vertical resolution for 480 active lines--things should
look fine on an interlaced set. If the vertical filtering is set as it
should be for a progressive scan camera at 90% of the potential vertical
resolution for 480 active lines, we may have problems when these images
are presented on an interlaced display. This is a real problem today with
film-to-tape transfers, especially when backgrounds contain fine patterns,
venetian blinds, etc.
If we feed the encoder with full vertical resolution, there are two advantages.
1. The improved detail actually helps the MPEG-2 encoding process,
providing improved compression efficiency (the net result is a similar bit
rate with more detail in the encoded pictures).
2. Progressive scan displays will be able to take advantage of the extra
vertical detail and make sharper pictures.
If we feed the channel with 480P at full vertical detail, it is still
possible to use an inexpensive convolution filter on STB's that feed
interlaced receivers. The filter will eliminate the small area flicker
artifacts caused by the extra vertical detail.
Regards
Craig Birkmaier
Pcube Labs
--
Craig Birkmaier
PCUBE Labs
Exploring the Convergence of Video, Computing and Telecommunications
Craig Birkmaier
In article <19970228152...@ladder02.news.aol.com>,
artal...@aol.com wrote:
> So you suggest that we not use the current HDTV equipment that is now in
second or third generation (analog and interlace) but instead wait until
someone builds the progressive equipment in a few years?
Jumping in here. The fact that interlaced HDTV equipment exists, even in
2nd and third generation products, does not make it's use practical or
affordable. The real test will be what products (formats) work well with
the underlying compression technology, and deliverr the best bang for the
buck. Clearly, even in third generation form, HDTV acquisition equipment
has not proven that it can deliver the required price versus performance,
and there is certainly no evidence that affordable display technology
exists to drive HDTV into the mass markets.
Perhaps now that Japan has awakened to the reality that 1125/60 and MUSE
broadcasts are not the best utilization of valuable satellite
transponders...now that they have recognized that "being digital" is more
important than promoting an obsolete analog HDTV technology, we will start
to see the real digital video acquisition and display products emerge. Mr.
Carliner was correct that progressive scan is the right way to go with
digital. There will be ample evidence of this at NAB, including
demonstrations of DVCPro capturing 480P at 60 frames per second.
> Who do you think should set this requirement?
The marketplace WILL decide. One big problem is that broadcasters are now
"the tail wagging the dog," in terms of what defines the marketplace. It
is ludicrous that they should define the future of video acquisition and
distribution, when they represent only a small percentage of total
equipment sales. Fortunately for broadcasters, they are already benefiting
from the emergence of affordable compression based digital video products
such as DVCPro. This time broadcasters will be able to leverage the
technologies that the marketplace will adopt, long before DTV broadcasting
reaches even 10% penetration.
> Production with modified HDTV equipment and conversion to digital before
broadcast seems to offer a much more rapid path to providing new HDTV
programing for the public. The progressive HDTV technology will compete
on a cost/performance basis and survive or not in the free marketplace.
Can you offer a reason why it should be otherwise?
The only industry with a valid reason to utilize HDTV production
technology today is the film industry. Because they are already acquiring
images with more resolution than the 2 million pixel HDTV formats, it
makes sense to transfer to the 24 Hz version of the 1920 x 1080 digital
format, or something closely related like 2K x 1K. This enhances the
archival value of the film, and can easily be re-encoded, at increasing
levels of resolution, as the marketplace evolves. Initially, from what I
have seen, Hollywood has no plans to encode for general release using
anything other than Main Profile at Main Level. There will be a few
programs encoded for HDTV distribution, because there are a few customers
who will pay the toll to prove to Washington that they are sincere about
HDTV--but the vast majority of content will be encoded for the resolution
that the masses can display, and this will not be HDTV.
As for local broadcasters, the idea that they will deploy 1125/60
interlaced acquisition and production equipment can only be considered a
joke. The real question is if NAB's constituents will ever do anything
better than MP@ML. I would content that this will only happen when they
start listening to organizations that want to help them become part of the
emerging digital world.
manf...@arl.bna.boeing.com
In article <craig-04039...@dialup-a1.lw.net>,
cr...@pcube.com (Craig Birkmaier) wrote:
[ ... ]
> Perhaps now that Japan has awakened to the reality that 1125/60 and MUSE
> broadcasts are not the best utilization of valuable satellite
> transponders...now that they have recognized that "being digital" is more
> important than promoting an obsolete analog HDTV technology, we will start
> to see the real digital video acquisition and display products emerge.
Maybe so.
> Mr.
> Carliner was correct that progressive scan is the right way to go with
> digital.
I have never seen anyone dispute this point, as a goal. No one. I _have_
seen people who know broadcasting say that to ease the transition, some
amount of interlaced transmission would be beneficial.
Of course, if the source material is film, presumably at 24 fps, then
there are _no_ interlaced options in the HDTV spec, so progressive is
mandatory. This would apply, for example, to the transmission of any
archived-on-film old TV shows over DTV.
> > Who do you think should set this requirement?
>
> The marketplace WILL decide. One big problem is that broadcasters are now
> "the tail wagging the dog," in terms of what defines the marketplace. It
> is ludicrous that they should define the future of video acquisition and
> distribution, when they represent only a small percentage of total
> equipment sales.
Sometimes, most of the time, I wonder who it is that you consider to be
the credible source for strategic direction in DTV.
> The only industry with a valid reason to utilize HDTV production
> technology today is the film industry. Because they are already acquiring
> images with more resolution than the 2 million pixel HDTV formats, it
> makes sense to transfer to the 24 Hz version of the 1920 x 1080 digital
> format, or something closely related like 2K x 1K.
The TV industry shoots most of its prime time shows on 35mm film, does it
not? So are you calling them "the film industry" as well?
> There will be a few
> programs encoded for HDTV distribution, because there are a few customers
> who will pay the toll to prove to Washington that they are sincere about
> HDTV--but the vast majority of content will be encoded for the resolution
> that the masses can display, and this will not be HDTV.
You're not saying anything earthshaking. In the early days of the LP, LPs
were mostly used to package together several 78 records' worth of
material. With FM radio, early FM radios could barely manage frequency
extremes beyond what AM could handle (5000 Hz). Eventually, that changed.
With DTV, at first, before a lot of high resolution sets are available,
maybe you're right. But as soon as any sets capable of better than 480 at
60P go for sale, broadcasters will want to attract those viewers.
And anyone with a screen that's bigger than 25" or so will notice.
> The real question is if NAB's constituents will ever do anything
> better than MP@ML. I would content that this will only happen when they
> start listening to organizations that want to help them become part of the
> emerging digital world.
Who are you talking about? It sounds like you're talking about religious
orders with nothing invested but words.