> God, I never thought I'd end up defending video, but there are just
> some pretty basic facts that are getting glossed over.
I agree, I didn't want to go into an detailed technical description
of the specifications of video and film standards.
> First of all, resolution is measured in *horizontal* lines, not
> horizontal and vertical.
Yes, in broadcast the resolution is measured in horizontal lines,
however I was thinking of an example which I saw last May where the
resolution was measured by a pixel count of a specially enhanced HDTV
image on a computer generated display. At the last International Council
of Archives Joint Technical Symposium the BBC and the East German Film
Archives cooperated in an experiment to see if HDTV and digital signal
processing could be used in the restoration of 35mm film. The Germans
used traditional photographic techiques on a reel of a 1942 feature shot
in colour. The BBC transfered the negative to digital HDTV tape and used
DSP to restore the film. We had an opportunity to examine both the
finished products. It was instantly apparent which was the video and
which was the film. The representatives from Ampex, Sony, BASF, and the
television engineers all agreed that even the best HDTV efforts in the
labs could not compete with the quality of film.
> Second, 1250 line is one of many HDTV formats, and
> there are some with higher and some with
> lower resolution. The only HDTV standard with production equipment
> available is an 1125 line system.
I am aware of some 15 different HDTV systems in development, The
system the BBC used in the experiment was a 1250 line system based on
a PAL compatible standard. Both the BBC and German Television are
working towards this type of standard for future production. I know
the the HDTV programs shot by the CBC and the Japanese MUSE system are
1125 lines due to the NTSC background of their existing systems. As
far as I am aware the only country with a regularly operating HDTV
system is Japan. Although programs shot with HDTV equipment have been
shown, in NTSC form on Canadian and US television. The Europeans have
also experimented with HDTV images broadcast in both SECAM and PAL.
The quality of the images as received in the consumers homes was only
marginally better than the normal broadcast image due to the limitations
of the existing broadcast chain.
> Third, Kodak-conducted tests indicate that the
> resolution of 35MM 5247 negative is comparable to a scanned image of
> 2330 lines (Mathias and Patterson, 231).
I have seen the Kodak test results and the images which they used to come
to their conclusions. They used a series of degraded 35mm transparencies
to simulate the scan lines of a single frame. The judgements of quality
were entirely subjective and were based on a still image. The 2330 line
image was the image that came closest to revealing all the image. I
found it interesting to examine the powerplant in the background through
the series as the number of smokestacks changed as the resolution
increased. At 1125 line resolution there was still one smokestack
> NHK tests indicate that the human eye cannot perceive a difference
> in resolution above 1600 lines (231)--under the
> conditions of the test, which means super-ideal.
Again the NHK tests were entirely subjective and based on a television
sized image of about 1 metre width. When you scale up the image a 1600
line image no longer is sharp enough. Under home viewing conditions a
1600 line image is probably "good enough" for most viewers. I suppose
that I can't be classified as a typical viewer.
> Still, the question remains as to whether the technology
> reaches a certain point where its simply overkill.
I firmly believe that different technologies have their niches in which
they are best suited. Current broadcast technology produces, in my
opinion, a marginally adequate image. HDTV, as currently proposed, is
considerably better. But HDTV does not come close to existing film,
regardless of its aspect ratio. The signal information is just not there
and will require at least one more generation of development beyond HDTV
to adequately approach the density of information a film can provide. As
to the question of overkill, that really depends on the applications to
which the technology is put. I don't think the afternoon soaps require
HDTV let alone film quality, however, a photomicrographic project will
almost certainly require film.
> I agree there are stylistic differences between video and film, but the
> implication here is that they are somehow innate to the technology.
> These differences are culturally specific. While they have a context
> arising out of certain technological factors, there is nothing about
> the technology itself that dictates television to be a close-up medium,
> or film to be panoramic one.
> We have ascribed these differences to the media, without recognizing
> our own expectations, by simply remembering what we have already
> seen and where we have seen it.
> According to your argument, a person trained in video could not shoot
> in HDTV, since the aspect ratio of HDTV is 5:2.
> Here are really two issues here; one involving composition that
> doesn't have anything to do with the resolution of video and the
> resolution of film,
> and the resolution of a 70MM print like Lawrence of Arabia. Under
> current standards, video would not compare favorably to such a print.
> But that doesn't mean that a similar effect can't be created with a
> similar aspect ratio.
I did not mean to imply that a person trained in conventional video
could not shoot in a widescreen format. I was pointing out that people
who have tried using video composition techniques for the "big screen"
have generally not been successful. Partly due to the scale of the image
and partly due to the limitations that a video producer must live with
due to the technology of video. This does not mean they cannot learn.
LAWRENCE OF ARABIA is a good example to use to compare the effect of
resolution on an image. Compare the VHS, Videodisc and film versions of
the scene where Lawrence is at the waterhole and he sees a small figure
approaching from the horizon. The figure just isn't there for most of
the scene on video. Of course, as you point out, current video
technology can't compete.
Mark Ritchie | Tel: (519) 888-4070
Media Librarian | Fax: (519) 888-6197
Audio-Visual Centre |
University of Waterloo | NetNorth: [log in to unmask]