Recording - December 2013 - 49

To explain this, I must back up and briefly
explain data compression. Until recently, fullresolution digital audio files were too large to
exchange on the Internet or store in a large
iTunes-like media library. Users therefore
needed to data-compress their files in order to
save bandwidth and/or disc space. Data
compression, which is not the same thing as
the dynamic compression discussed in this
article, removes masked information from
audio files. Using complicated psychoacoustics, data compression codecs (coderdecoders) remove auditory information that
louder sounds in a recording cover up. Quiet,
subtle, and background details are erased for
the sake of making the file smaller.
Data compression has long been a contentious issue in our industry and thus various codecs and their parameters see regular discussion and debate. The issue at hand
here is that hyper-limited, constantly loud
audio files offer little in the way of low-level
background details to be removed by a
codec. Modern country, rock, hip-hop, and
electronic dance music tracks are often thickly layered with overdubs, and then
squashed to the max to boot. All frequencies
are essentially loud at all times. Because of
this, codecs can be less effective and errors
and distortion can be far more prevalent.
Creeping loudness
Since the late 1990s the loudness wars
have escalated, with many music producers adopting new limiting technologies, not
to lower sonic artifacts, but to push levels
even higher. For the most part, the general
public remained unaware of the issue until
2008, when Metallica's Death Magnetic
album was released simultaneously with a
companion Guitar Hero video game. A
group of fans, having bought both the CD
and game, discovered that the album version was louder, darker, and more distorted. Some fans, angered at the CD's poor
sound quality, demanded a refund; others
(and the band) argued that it sounded just
fine. The media took notice and the loudness wars became international news.
In recent years, following this debacle
and the growing awareness of loudness
issues, the music production community has
begun to acknowledge that louder is not
necessarily better. The ugly truth is that
while an overly compressed recording may
initially pop out of the speakers when
played back-to-back with quieter material,
the listener usually responds by simply turning down the playback level. The net result
is that the loud recording is once again soft,
but missing dynamics at the expense of
fidelity. Oddly, this is the very scenario that
may eventually end the loudness wars. In
Europe, regulations now require that radio
and TV broadcasts be "loudness normalized" so all material is of equal loudness.

American television is also regulated in a
similar manner. With regulations like these
becoming commonplace, is it worth it to
trade fidelity for loudness?
Similar to what has been adopted in
the broadcast community, Apple's iTunes
program offers a loudness normalization
feature called "Sound Check." When
engaged, iTunes analyzes its library and
imbeds average loudness metadata in
each song file. Then, during playback,
each song is adjusted for consistent loudness. With Sound Check on, gently compressed and overcompressed files play
back equally loud, with crushed recordings simply sounding worse.
Wrapping things up, I recommend a few
other resources for better understanding of
loudness and compression. The first is
iTunes Music, a book from renowned mastering engineer Bob Katz. Affordable and
easy to read, it will enlighten any engineer
to the realities of producing music in the
digital era. Another invaluable resource is
the great collection of loudness-associated
videos available on YouTube. Matt
Mayfield's "The Loudness War" is probably
the best-known clip, and it eloquently
demonstrates how an overcompressed
recording sounds lifeless and crappy when
returned in playback level and compared
to its uncompressed self. The video might
just change the way you mix!
Even with three months of focusing on
compression, our investigation here has
only been cursory, though, as the topic
could feed a lifetime of study. Once
again, the best teacher is your own
ears-experiment with these techniques,
good and bad, and learn how they
sound.
Goodbye for a while...
For the last twenty-four months I've examined one attribute of audio-geekery after
another, and it's time for a much-needed
[and well-deserved!-MM] break. So with
this chapter wrapped up, I'll be going on
hiatus for a while.
However, Recording Fundamentals will
return-with occasional articles exploring more complicated topics than are
suitable for this monthly format. Thanks
for reading, and I look forward to bringing you more Recording Fundamentals in
the future!
Eric Ferguson (ferguson@recordingmag
.com) spent a dozen years as a freelance
audio engineer in Los Angeles, and is now
on the faculty of the New England School
of Communications (NESCOM) in Bangor,
ME. The Editors thank Eric for his tireless
contributions, and we look forward to
bringing you more... as many as Eric's rigorous teaching schedule will allow.

http://www.recordingmag.com

Recording - December 2013

Table of Contents for the Digital Edition of Recording - December 2013