| Video Noise December 2007
 What's So Different About Digital Video?
Just the other day, our publisher sent me an e-mail asking
a simple question: "Are you seeing any differences between HDMI cables?" I
suspect he may have already formed his own answer to this question, and wanted to see if
anyone else had been paying attention. I replied that, no, in terms of video, I see no
differences between High-Definition Multimedia Interface (HDMI) cables.
I began experimenting with HDMI connections in October
2006, when I got a new 1080p display with three HDMI inputs. Since then I’ve
collected a number of HDMI cables: a free one that came with one of the components I
bought, under-$5 and under-$17 models from www.monoprice.com, a $50 Acoustic Research cable, and a $135 Belkin.
They all perform identically. When I change cables, I see no difference, and I can’t
measure differences in color performance even using professional tools.
This is a very different situation from the days when the
cathode-ray tube (CRT), an analog display technology, ruled the world, and every video
cable produced a different-looking image. In those days, everything mattered. Power
conditioners almost always improved images. Different cables helped or hindered them. Even
shifting a large-screen set’s position in the room could change its image quality --
the electron beams traveled such long distances within the tube itself that the
Earth’s magnetic field could slightly deflect the beams of electrons, and thus
slightly alter the image.
Digital video is different. I’ve tried seven different
power-conditioning products with my LCoS digital display, and only one of them, a balanced
model, made any detectable change in image quality -- but the change was so tiny that
I’d hesitate to recommend that anyone spend $3500 on that conditioner. None of the
passive power conditioners made any difference at all.
Why could, or should, that be? Digital video is a series of
still images created by ones and zeros transmitted through wires; usually, 30 images per
second. You don’t actually see motion; your brain creates the illusion of motion from
the progression of still images, much as in animation. Each 1080 still image transmitted
via an HDMI or DVI cable is built of more than 6 million numbers. Repeat that delivery 30
times per second, and that’s a lot of numbers. Each number tells one of the
display’s pixels to create a specific color of light, and all of those pixels flash
at the same time. In a CRT, a single beam of electrons scans the back of the screen many
times to create an image one pixel at a time. A digital video display flashes the entire
still-image frame on the screen for 1/30 second, then replaces those pixels with a
completely new set of pixels for another 1/30 second. There is no scanning, and nothing to
influence the pixels to be anything but "perfect," based on the numbers coming
through the cable.
Still, power conditioning remains important to digital
video, if only for nonvisible reasons. You want to protect your equipment from damage due
to surges and spikes. You want the voltage to the lamp or other illumination system to be
as steady as possible, so that the image brightness remains stable. And you may want to
protect the lamp and optics from heat buildup in the event of a power failure by having a
backup battery power-conditioning system that produces a good sinewave, so that the
cooling fan can run for a few more minutes. But it looks as if the days of using power
conditioners to get better images may be over, at least for LCD and DLP displays. I
haven’t yet experimented with plasmas, whose technology is different enough that it
may be more responsive to power conditioning than other display technologies.
 Two parameters of the HDMI protocol that
might make a more expensive cable a better buy than a less expensive cable are a
cable’s length and its bandwidth. HDMI cable length is such a quagmire that HDMI
Licensing, LLC (www.hdmi.org), the
consortium of manufacturers that sets HDMI specifications, won’t make even vague
statements about the maximum length of cable you can use between components. There are
just too many variables, including the transmit/receive chips in your components and the
quality of the cable itself. Inexpensive cables might work fine in lengths of up to
15’ or 20’ before any visible problems occur. The most common such problem,
"sparklies," makes random individual pixels shine more brightly. But increase
the length of such a cable and you could lose the image altogether. However, an HDMI cable
of higher quality might be OK up to 35’ or so.
Regarding bandwidth, the HDMI 1.3a specification supports
the transmission of more video data plus a lot of audio data, including uncompressed
audio, which requires more bandwidth; i.e., higher frequencies. Although it
won’t be an issue for a while -- at present, no sources use the full bandwidth
supported by HDMI 1.3a -- at some point in the future you may find that your old HDMI
cables are no longer up to the job.
In a future column, I’ll look into another aspect of
HDMI cables: digital audio. Digital audio signals aren’t buffered and displayed one
frame at a time, like digital video signals; they’re converted to analog signals in
real time. So digital audio remains sensitive to influences by cables and power
conditioning.
Next time you see a 2m HDMI cable priced at $100,
you’ll know that there’s no reason to spend that kind of money -- it won’t
buy you any increases in image quality. But will the sound be reason enough to
justify an expensive HDMI cable? It might take me two or three months to get back to you
with an answer, but stay tuned.
...Doug Blackburn
db@hometheatersound.com |
