Brian Doherty | October 4, 2006
Ever wonder why a modern Christina Aguilera or Red Hot Chili Peppers CD can hurt to listen to--even if you dig the music? Check out this marvelous contribution by Joe Gross at the Austin-American Statesman to one of my favorite tho alas rare genres of popular journalism, audiophile-philia (and phobia), for explanations of why, as Bob Dylan insisted, all modern recorded music sounds annoying and awful.
The key--read the whole article if you want the deep stuff, if anyone still does in these days when MP3s seem to satisfy most of our music listening needs and the days of even semi-mass popularity for multi-thousand dollar turntables, speakers, and special plastic discuses stuck on your walls for 1000 percent fantasy audiophile ultramegaperfection seem like a weird dream of a very wealthy world's amusingly mad indulgences--is here, starting with a quote from a:
letter...written by Angelo Montrone, a vice president for A&R (the folks who scout and sign music acts) for One Haven Music, a Sony Music company.
"There's something . . . sinister in audio that is causing our listeners fatigue and even pain while trying to enjoy their favorite music. It has been propagated by A&R departments for the last eight years: The complete abuse of compression in mastering (forced on the mastering engineers against their will and better judgment)."
This compression thing has been a topic of discussion among audiophiles and music fans for nearly a decade. But hearing a music industry executive cop to it was pretty unusual.
..............
"The mistaken belief that a 'super loud' record will sound better and magically turn a song into a hit has caused most major label releases in the past eight years to be an aural assault on the listener," Montrone's letter continued. "Have you ever heard one of those test tones on TV when the station is off the air? Notice how it becomes painfully annoying in a very short time? That's essentially what you do to a song when you super compress it. You eliminate all dynamics."For those already confused, Montrone was essentially saying that there are millions of copies of CDs being released that are physically exhausting listeners, most of whom probably don't know why their ears and brains are feeling worn out.
I will note that as an inveterate and irredeemable CD shuffler, that Dylan's own latest-not-greatest, Modern Times, sounds waaaay too loud at the same volume as any pre-2000 CD I shuffle it with. (This phenomenon discussed deep in the article.)
I stole this post title straight from Gross' article, as it was too perfect.
Help Reason celebrate its next 40 years. Donate Now!
Try Reason's award-winning print edition today! Your first issue is FREE if you are not completely satisfied.
Is this news to some people? The complaints about overcompression have been, um, quite loud for years.
So let's get this straight - compression, which has been around
for dog's years, has only been heavily used on pop records starting
around 1998? Just 'cause? I'm no audio engineer, but I smell (or
hear) a rat. Some of the stuff in the article just doesn't quite
jibe for me, like this:
We're looking at the wave forms generated by a number of modern albums. Sound waves should look like what they're called: waves, with sharp peaks and valleys. But the music we're looking at is all peak. It's like looking at a butte or a brick.
"These square waves are a very unnatural occurrence," Wofford says. "It sounds wrong to the ear. You can't hear detail."
If everything is squared off, that's clipping, not compression -
and they mention clipping, but the author gives you the impression
they're the same thing, which they're not (and which the recording
engineers sure as hell know). You can have something all compressed
to hell and back, but not clipping.
I'm impressed by the large number of double-blind studies they cite
to prove that newer recordings really are perceived as
qualitatively unpleasant by the majority of consumers - oops, wait,
they didn't cite anything. Yes, clipping and overcompression sound
bad, but they haven't done anything to show that people actually
think newer recordings sound bad because of this. So what are they
but a buncha audiophools pushing their pet theory to "prove" once
again that all the old stuff was better?
"If everything is squared off, that's clipping, not
compression..."
It's "clipping" if the peaks exceed the maximum capabilities of a
given medium. (In the digital realm, that maximum is 0 db.)
A waveform that clips can appear square, but it does not
necessarily appear so.
A heavily compressed waveform, on the other hand, necessarily looks
square, because the troughs now match the peaks.
I'm all too familiar with excess compression - from my days as
an AM Top 40 disc jockey. The idea then was to make the station so
loud it "jumped out" at you when you tuned by. In reality it gave a
lot of those old stations a "throbby" quality that would give you a
headache after a couple of hours listening.
Modern music stations generally don't use as much compression
(perhaps because the music itself is often so highly compressed).
Today's record-buying public is much more audio-savvy; I see a
backlash coming.
Also: compression on music is nothing new; Phil Spector and Berry Gordy used the technique in the sixties to make their recordings stand out more on the crappy little two-inch AM speakers of the era. That's is why if you ever tried to play an old Motown single (say, the original pressing of Reach Out, I'll Be There by the Four Tops) on a newer high fidelity stereo system, it sounded awful (this mainly applies to original pressings through the end of the sixties; later issues, and repressings of the originals are mixed and processed to be more fidelity-friendly).
I'll just complain that my New Pornographers CDs are shriller
than Cindy Sheehan.
Dunno if that's from compression, or from cranking up the treble in
the studio, but it gets tiresome.
Rick Rubin produced recordings, to my possibly mistaken ears, sound very compressed, but, on the Johnny Cash American Recordings in particular, this seems to be part of their charm. I get a feeling listening to them that Johnny is about two feet away from me. Is this the effect of compression?
It's "clipping" if the peaks exceed the maximum capabilities
of a given medium. (In the digital realm, that maximum is 0
db.)
A waveform that clips can appear square, but it does not
necessarily appear so.
A heavily compressed waveform, on the other hand, necessarily
looks square, because the troughs now match the peaks.
Exactly right.
I'm impressed by the large number of double-blind studies they
cite to prove that newer recordings really are perceived as
qualitatively unpleasant by the majority of consumers - oops, wait,
they didn't cite anything.
You have a point, but the fact that the majority of consumers
aren't particularly savvy to sound quality and dynamics doesn't
really impact the central claim that overcompression is a big
problem in the modern mastering process. It's true though, that
they try to sell this as "overcompression is the dirty little
secret explaining why much of modern audio sounds off-putting to
everyone" when really most people don't have the musical ear to
fully appreciate the difference.
As Tom mentioned, over-compression is something audiophiles have
noticed and complained about for years, whereas the "majority of
consumers", who are unlikely to sit down and listen to Serge
Gainsbourg's "Histoire de Melody Nelson" and Deftones "White Pony"
(a prototypical "hot copy" mastering job) side-by-side for
comparison purposes, are also unlikely to notice the differences
between music recordings then and music recordings now. The claim
that said people experience "fatigue" and "pain" is unsubstatiated
and perhaps a bit hyperbolic, but if this article draws more
people's attention to this stupid practice and even puts an end to
it, I'll forgive them their rhetorical flourish.
The A&R guy is not so much complaining about compression
itself as much as master buss compression, where the two (or six or
seven, in a 5:1 or 6:1 mix) left-right stereo tracks are squashed
to death in mastering. Compression is usually best reserved for
individual tracks, and on a widely varying basis for each. It's
common to avoid using it on individual tracks that need greater
dynamics, but you might use it to even out a bass part or a kick
drum or a singer who doesn't know what they're doing.
But there is indeed truth in what he says as it refers to mastering
a CD. While it's not really scientific to judge one twenty year old
CD's weak overall level to one made today (as analog-to-digital
equipment has vastly improved in that span), you will notice a huge
difference in the mastering levels of older vs. newer discs if you
reference, say, the original CD you bought twenty years ago of
Steely Dan's "The Royal Scam" to the more recent Remaster. The
audiophile-quailty is still all there, and it's a helluva lot
louder, because the mastering engineer used new fancy tube analog
equipment to bring the level up (before converting to digital)
without using buss-level compression. That's the effect you're
hearing on the American Recordings stuff.
The problem comes in with more modern music, where the stereo
master track is all shoved into a compressor at the mastering
stage, and "squashed". I can mainly attribute this to Nu-Metal
garbage like Limp Bizkit and Korn for making this a trend about ten
years ago (Tool's "Aenema", by contrast, is still a richly dynamic
record), but I agree that the blame lies mostly with record execs
who interfere with the mastering process. Engineers know that ears
tire with too much of that shit, but executives want to play God
with your volume knob so you won't have a choice but to hear it,
and then BUY, BUY, BUY!
I don't think the trend is likely to subside in the recent future
as downloaded music picks up market share, since they will feel the
need to control iPod volume in the same way.
Everything Louder Than Everything Else
Hey! That's one of my favorite Meat Loaf songs! Nice touch
Brian.
some of this is making things waay too loud.
and some of it is merely aging, and changing musical styles, i
think. there are plenty of artists which rely on blasting the
living shit out of stuff. which in turn is blated out of the
windows of cars, so who knows. hip hop made everything bump a
little harder in the pop world, and stuff has to fit over the
bass.
hip hop made everything bump a little harder in the pop
world, and stuff has to fit over the bass.
Actually, you should blame or credit disco. Disco brought the drums
and bass up front. Listen to rock pre & post disco by the same
artist and you will hear the difference.
I have noticed that some albums certainly do wear out my own
ears very quickly. When I was younger, I'd bicycle to Eliot Smith's
Figure 8, and I couldn't do it for more than a half hour or so
before it gives me a splitting headache. (The New Pornographers, as
someone mentioned above, have a similar effect--it's a shame, as
Eliot Smith and TNP both put out durn fine records.)
I'm not sure if it's a result of compression, or just my ears being
overly sensitive to treble levels, but I certainly don't notice it
on older albums.
Technology to the rescue! You need to start encoding your MP3s
with ReplayGain:
http://www.replaygain.org/
Or, better yet, your FLACs. Which you listen to with Foobar2000 on
your computer... and on the go using your RockBox-hacked iRiver...
or maybe that's just me.
larry
Ironically, one of the original selling points for the CD format
was that its broad dynamic range and inherent noiselessness would
allow record companies to abandon the compressors and peak limiters
they were having to use when mastering for vinyl or cassette.
"Properly recorded and mastered," we were told, "CD sound could be
indistinguishible from that of a live performance," whatever the
musical genre.
Let's not even count that some people say that they can hear
"brittleness" or other objectionable qualities in even the most
carefully produced CDs. The claim of pristine sound via CD was a
crock from the beginning, because anyone who had spent any time
around the radio business or music industry knew that economic
pressures would eventually be exerted on producers and studio
engineers to "tweak" the sound in various market-friendly
ways.
No technological solution is greed-proof.
Larry Hastings,
What difference does the file format make if the recordings were
mastered a certain way? Don't the source files all share the same
mastering?
http://www.replaygain.org/ ?
Sounds like it just controls the volume of each track, but doesn't
address compression.
You can un-do the compression, to a large extent, with an 'audio
expander.' Like this one:
http://www.cakewalk.com/Products/AudioFX1/Expander.asp
I'm impressed by the large number of double-blind studies they
cite ... oops, wait, they didn't cite anything.
Heh - ABX testing would put most of the audio industry out of
business! (Because people can't tell the difference between a $30
Radio Shack CD player and a $5,000 botique CD player...*if* they
don't know which one they're listening too)
A heavily compressed waveform, on the other hand, necessarily
looks square, because the troughs now match the peaks.
No, not really.
Derek Ashworth has it exactly right. I'm a musician who records
my own stuff, and I've studied the art of recording for many years.
All of the ridiculous mastering-stage compression these days is
intended 100% to make sure that the song is as loud or louder on
the radio than everything else. Which is kinda silly, since radio
signals are highly compressed even further.
Interestingly, Iron Maiden decided not to put their new cd through
the mastering process at all. The cd you buy is a straight transfer
of their final mix. Personally, I'm not a big fan of their new cd
from a musical standpoint, but that's a pretty bold step for a
major-label artist to take.
A lot of this is the product of the cd remastering craze of the
mid-late '90s. The better remasters were done very carefully to get
both higher volume and more faithful reproduction than the original
cd versions, which were often mastered from inferior sources, such
as "production masters" used to make LPs. But with thousands of
titles to remaster and sell to the public yet again, the major
labels often just took the same sound source, compressed the shit
out of it, and passed it off as a 'remaster.'
By the way, putting an album through an audio expander might undo
some of the problems with over-compression, but it won't undo all
of the digital distortion that over-compression creates, which is
one reason so many new recordings sound so harsh. Analog
compression would lead to clipping, but the result could often be
musically useful, as with getting a good snare drum sound, but
digital distortion is just ugly.
There is no reason to believe at an overcompressed music source
would create any challenges for the auditory system that would lead
to headache or "physically exhausting listeners." There is an
auditory reflex which protects your auditory system from loud
sounds that tires with continuous exposue, but this has more to do
with absolute volume than dynamic envelope.
Any complaints about this come from aesthetics... not
neurocognitive fatigue.
Right, all ReplayGain is doing is automatically adjusting the
volume level of your player to maintain a generally consistent
volume level. It's not de-compressing the sound, which I imagine is
as likely as de-compressing a crushed car. But I find that my ears
are less fatigued when I listen using ReplayGain, which is why I
recommend it.
larry
Overcompression isn't anything new, but anyone who cared used to be able to dial the midrange down a bit. But now stuff is SO overcompressed you can't turn down the midrange enough.
"I stole this post title straight from Gross' article, as it was
too perfect."
And Gross stole if from
Deep Purple..."Made In Japan" (instructions given to sound guy
heard between songs...)
Here's a wiki going into some other detail about the
subject:
http://en.wikipedia.org/wiki/Loudness_war
Also...
Deep Purple's "Made In Japan" can be used an example of the
practice; the MFSL CD version is a much different listening
experience than the Rhino remastered version.
Site comments/questions:
Media Inquiries and Reprint Permissions:
(310) 367-6109
Editorial & Production Offices:
3415 S. Sepulveda Blvd.
Suite 400
Los Angeles, CA 90034
(310) 391-2245