Loud TV Commercials a Matter of Signal Processing
Few things are more irritating when you're watching the
tube than a commercial that seems to jump out and smack you in the face. It's not just the
ones where some obnoxious twit is shrieking at you to buy appliances for the sixteenth
time that evening -- even ads that are not so inherently offensive often seem much louder
than the programs they interrupt.
Most people assume that the TV stations play the
commercials louder to make us pay attention to them; they are, after all, the
broadcasters' bread and butter. But all TV types deny that there is some guy in the
control room saying "Here comes a commercial, let's crank it up a coupla
notches." And yet, everyone has heard these level differences and been annoyed by
them to a greater or lesser degree.
It's not your imagination, but neither is it a conspiracy
on the part of the TV stations to give some extra prominence to their advertisers.
To understand what's going on, you have to know something
about the nature of an audio signal and how we perceive it. Sound consists of a series of
air pressure differences created by a vibrating object, which travel through the air until
they reach our eardrums. The pressure changes move the eardrums back and forth and we hear
the sound. For recording or broadcasting, a sort of artificial eardrum called a microphone
turns the pressure differences into electricity in the form of a type of alternating
current.
Unlike the AC that comes out of the wall plugs in your
home, which is perfectly smooth and regular, an audio signal is very jagged and irregular,
as is sound itself. The more complex the sound, the more irregular the audio waveform --
musical signals are particularly convoluted.
Unless they're doctored somehow, most audio signal share
one characteristic: most of the time the level is quite low, with only occasional very
brief peaks. These may be many times the volume of the material surrounding them, but as
they are normally of such short duration, they don't seem excessive.
The peaks are very important when making a recording or
when broadcasting, however. All recorders have an inherent level ceiling, beyond which the
system will produce unacceptable levels of distortion, so in order to make a
"clean" recording, the peaks have to be kept below that maximum level. In
broadcasting, it's also a matter of law: stations are not allowed to go beyond a certain
maximum level.
But although the audio peaks determine how high a sound
engineer can turn up the level, they have virtually nothing to do with perceived loudness.
That is determined by the much lower average level -- the level the signal is at
most of the time.
In most material the "peak-to-average ratio" is
pretty constant, and so two adjacent bits of audio will sound pretty much the same in
terms of loudness, unless one of them has been electronically altered.
Such processing is pretty common in broadcasting, it turns
out. Sophisticated devices exist that either compress the signal -- bring up the low
levels and reduce the high ones -- or simply limit the peaks to a predetermined level.
Both have the effect of reducing the peak-to-average ratio, which means the engineer can
turn the signal up without exceeding the maximum permissible level. The result is a high
average level and a seemingly louder sound.
The use of such techniques varies widely. AM radio often
squeezes the signal so much that it has a dynamic range of only a few decibels. Try
recording a stretch of AM programming and watch the level meter on your tape deck; you'll
be amazed at how little it moves.
FM typically relies much less on compression, and
television programmers even less. But not the makers of TV commercials. These are often
severely compressed so that if the audio guy at the TV station sets his level so the peaks
are right, the overall sound will seem considerably louder.
...Ian G. Masters
ian@mastersonaudio.com
|
