Issue 222 April 2012
Losing Half The Music: Why Audio Sounds Wrong
Editorial By Robert E. Greene
Musicians are interested in, even
obsessed with, sound. Thus it seems paradoxical that musicians are typically
not very interested in audio. But in my experience, musicians are fascinated
with audio if it sounds right. The trouble is that most audio systems sound
fundamentally wrong. This is because most audio systems have a big hole in the
frequency range just below middle C, and to someone who listens to sound as
music, not in audio categories, this is a fatal flaw.
Middle C on the piano is 261.6Hz. Middle C is the natural
dividing line between the upper and lower ranges of music, the boundary in
particular between the treble clef and the bass clef. Pick up a hymnbook or
any other place where four-part vocal harmony is written. The basses and the
tenors are in the bass clef, the altos and sopranos in the treble. Similarly,
the left hand of the piano usually plays the notes below middle C and the
right hand those above. Of course, the ranges overlap. Still, closely enough,
half of music is below middle C, half is above.
But a startling percentage of the speaker/room combinations
in the real world have a big hole in the range below middle C. This is far
more common than not. Somewhere between 100 and 300Hz, there is typically a
deep and quite wide trough in the in-room frequency response. The problem is
not primarily the speakers' anechoic response, but the speakers' not being
designed to interact with the floor correctly. Some of the designers who
presumably cannot figure out how to deal with this contend that it is not a
problem and that floor bounce is a natural thing. This is nonsense. The
natural floor interaction of the original event is already recorded, and
whatever floor interaction a system adds, if it does, is spurious — and
I am not speculating about the existence of this dip. It is
easily audible and it is easily measured. And in this frequency range, what is
measured is what is heard.
And you can find many more examples any place where in-room
response measurements are available. There is a good reason why Martin Colloms
in HiFi News used to call this
dip "the usual floor dip" — it really is usual, especially for
It is also enormously destructive to music. If energy is
lost in the octave below middle C, concert grand pianos turn into spinets and
the cello section goes out to the lobby. And the harmonic foundations of music
are undermined, and much of its power and grandeur and beauty is destroyed. In
addition, because deep bass instruments are heard in good part through their
harmonics, the deep bass also loses perceived energy, compounding the damage.
Having the right balance through this lower midrange and
upper bass region makes all the difference. The world is all too full of
speakers that sound thin and tinny because of that floor dip. It is easy
enough to discover if your system suffers from this problem. If you do not
trust your ears — and you might be too used to this wrong sound to react to
it as strongly as it deserves — just try measuring with test tones and an
SPL meter. A rough measurement, but it will do the job because the effect
involved is large, typically 3 to 5dB or more.
I have done a lot of DSP room/speaker measurement and
correction of a great many systems over more than fifteen years. Take it from
me, by far the most common and most troubling problem of speakers in rooms is
precisely that the crucial octave-and-a-half below middle C is down in level
in part or entirely. Applying the appropriate correction makes a truly major
difference to music. No subtle matter here. With the correction, there is real
music. Without it, there is a thin, tinny imitation. This lack of energy below
middle C is typically the biggest error in listening terms that room/speaker
combinations make. And it ruins music.
If you want your system to sound like music, look for
speakers that interact with the floor correctly, and also experiment with DSP
room correction. Getting rid of the usual floor dip makes all the difference.
It really does.
-- Robert E. Greene
Click here to subscribe
to The Abso!ute Sound.