N
o digital media player can satisfy everyone. Even the exqui-
site iPod Nano is pissing off folks—justifiably—with a surface
so delicate it can be marred by harsh language. And the spacious
screens on high-end portable video players look ridiculous to some-
one who just wants a pocketable player that doesn’t require its own
duffel bag. What we can all agree on, however, is that sound and
video quality are paramount.
There’s no reason to whip out the heavy machinery to distinguish
one player from another—we’re dealing with compressed audio
here, folks. So after covering the basics (codec support, battery
life, viewing angle, etc.), I put on Shure’s E4c earbuds ($300, http://www.shure.com) and listen to five audio tracks and three videos of differ-
ent genres, and I repeat the tests for all supported codecs. If there’s
any doubt, I also perform audio A/B testing with a pair of studio
monitors. I’ve enjoyed most of these tracks for more than a decade,
and the music is so imprinted upon me that if a single hi-hat tap is
muted, I hear it. And that’s what’s important, because what might
seem to be a minor compression artifact could ruin the ragged com-
plexity of a distorted guitar, or smear the nuances of a tasty Chopin
Nocturne. The same can be said about fast-motion and deep-black
reproduction in video players. Conveniently, chapters 12 and 13
from the Crouching Tiger, Hidden Dragon DVD have both elements,
and I use this video to check for blurring and blockiness across all
supported codecs.
Just keep in mind that no matter how good your MP3 or portable
video player is, it’s only as good as the encoding job you’ve done
compressing your media. Want to encode better? Check out the
forums at http://www.hydrogenaudio.org and http://www.vcdhelp.com to pick up
some tips from the truly obsessed.Logan Decker
ON WHAT IT TAKES TO TEST
MEDI A PLAYERS
Introducing the latest weapon in our videocard
benchmarking arsenalTrust your ears, trust your eyes
MA XIMUMPC DECEMBER 2005
iHthe F;< REAL-WORLD TESTING: RESULTS. ANALYSIS. RECOMMENDATIONS
MICHAEL BROWN
Explores the
HQV Benchmark
DVD
W
hen I measure a videocard’s performance, I use a variety
of games and games-oriented synthetic benchmarks
to test a card’s capacity for real-time 3D-rendering. But
because I know Maximum PC readers are much more than avid
game players, I’ve been looking for a benchmark that will enable me
to objectively measure a videocard’s video performance, too. The
HQV Benchmark DVD delivers just what I need.
Silicon Optix, in conjunction with several home-theater maga-
zines, developed this disc for measuring the performance of high-
definition televisions, DVD players, and video-processing gear, but
it’s proven to be an eye-opening benchmark for videocard testing,
too. (Silicon Optix, by the way, is a fabless semiconductor manufac-
turer specializing in video-processing technology for the consumer,
broadcast, and post-production television markets. Because the
company is not in the videocard business, I don’t need to worry that
its benchmarks are skewed toward any one GPU manufacturer.)
The HQV Benchmark DVD consists of 12 acid tests based
on real-world scenarios. It presents videocards with a very tough
MPEG-2-decoding obstacle course involving de-interlacing, motion
correction, noise reduction, film cadence detection, and detail
enhancement. I use the Color Bar/Vertical Detail test to evalu-
ate color saturation and the card’s capacity for vertical resolution.Several tests measure the videocard’s ability to remove scan-line
artifacts—such as jagged edges on diagonal lines—while de-inter-
lacing video. Other sequences help me analyze the videocard’s abil-
ity to display fine detail, accurately reduce visual noise, and properly
detect the cadences that have been used to convert film, animation,
and anime to video.
The disc provides explicit instructions on what to look for in each
video sequence. It also displays examples of what passing and fail-
ing sequences look like. Thanks to these characteristics, I’ve found
the test results to be repeatable and therefore reliable.
Right before we sent this issue to press, however, Silicon Optix
informed me that my HQV scores seemed too high. To resolve the
issue, Silicon Optix agreed to loan us a DVD player that uses the
company’s Realta HQV decoding engine: Denon’s $3,500 DVD-- According to Silicon Optix, the DVD-510 scores a perfect 130
and will serve as a reference point for our tests. I’ll let you know
the results in our next issue. Meanwhile, if you’d like to run these
tests on your own rig, you can purchase the $30 disc at http://www.hqv.
com/benchmark.
iHthe F;< REAL-WORLD TESTING: RESULTS. ANALYSIS. RECOMMENDATIONS
We use Silicon Optix’ HQV Benchmark DVD to torture-test a
videocard’s capacity for accurately decoding MPEG-2 video.