Slashdot Mirror
Does Going Digital Mean Missing Music?
arlanTLDR writes "The Seattle PI is running a story about how the MP3 format is the sign of a musical apocalypse. Apparently, many top music producers are 'howling' over the fact that files in a compressed format contain 'less than 10 percent of the original music on the CDs.' Is this just sensationalist FUD, or is there something to the assertion that listening to an MP3 is like hearing music 'through a screen door?'" The article mentions that the iPod and its cheap earbuds bear some of the responsibility for rendering this degradation in sound quality less objectionable.
I remember AM tube radios.
Now quit complaining and get off my lawn.
You will be surprised at just how much of that 90% of sounds produced our ear cannot understand.
Kevin Smith on Prince
It's just whining. There have been numerous double-blind ABX tests, many done by the folks over at Hydrogenaudio.org, comparing MP3 files to AIFFs, and with the right codec and right bitrates (depending on the type of source material), it's possible to get an MP3 that only the most refined ears can discriminate from the original. [1]
a ssical/mp3test.html. While well-trained ears could discriminate between 128kbit MP3s and PCM, they could not reliably tell the difference between 256kbit and PCM, on average. This is just the tip of the iceberg.
Of course, it's quite possible to make an MP3 that sounds like a tin-can telephone with one end held underwater, and I'd argue that many of the consumer-ripped files floating around the P2P networks fall into this category, but these files only exist *because* there aren't legitimate, professionally-made, DRM-free MP3s. (And because some people like getting stuff for free and don't much care about the quality when they do. But I do think there is a market for and profit in digitally-delivered music, for the people who can do it right.)
As more music begins to be distributed as MP3s, sound engineers will doubtless (if they have not already) begin studying the codecs and encoding procedures in order to wring the most quality out of a particular bit rate. Many amateurs and enthusiasts have already done this, and there is a sizable body of work devoted to the topic -- including the LAME encoder itself.
Also, looking towards the future, while CDs have pegged the standard for digital music as 2 channel, 44.1kHz, 16-bit PCM, there is no reason why an appropriately-crafted MP3 file cannot *exceed* it in terms of quality. The Apple iPod already supports (slightly) higher sample rates, I believe, and if consumers desire it [2], there's no reason why modern digital formats cannot encapsulate very high-definition audio.
The only people who I hear whining about MP3 are those with either an ulterior motive and a desire to try and keep the industry from moving away from a distribution model that revolves around physical objects, or those who just don't understand the technology. (There are a very small core of audiophiles and techies who seem to dislike MP3 because they prefer some other format, usually either for ethical/political reasons or technical ones, and there certainly is an argument in favor of using lossless formats in lieu of MP3 for distribution, but overall MP3 strikes a good balance between quality and portability. [3])
[1] One 'competition' that pitted serious self-described audiophiles against modern codecs is described in detail here: http://www.geocities.com/altbinariessoundsmusiccl
[2] Which is a big 'if.' The buying public, to date, has shown little interest in high-definition audio as such. The only exception to this is multichannel audio, but that only in movie soundtracks for surround sound.
[3] This does raise the question, though, of why the legitimate music-download sites don't take a cue from the late, great, AllOfMp3.com and just allow the *customers* to choose their format of choice for their downloads. There's really no particular excuse not to at least offer a few different quality/size options, particularly for popular music that is going to be enjoyed in a variety of settings (automobiles, portables, home stereos -- each lends itself to a slightly different EQ and compression).
"Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
While the 90% figure may be overblowing things a bit, there is a noticeable loss of sound fidelity when converting to a compressed format. In fact, it's actually quite impressive that the loss is not even more noticeable than it is, and that is a testament to the brilliance of the original MP3 algorithms, which have been tweaked and honed to make the quality even better.
The fact remains, however, that most listeners, in most situations, don't care. For one thing, popular music has, since the 50s, been designed for listening to on cheap equipment. The dynamic range is enormously compressed, the sounds are often fuzzy to begin with, the voices are straight front and center. This can explain the dwindling popularity of classical and jazz, and the rise of the louder, simpler, more beat-oriented music like rock, rap, or pop. Note that I'm not saying the music is of lower quality, but that it can be reproduced "faithfully enough" on lower quality equipment.
I don't have any statistics, but I would bet that most music listening happens while the listener is doing something else: driving, working out, coding python scripts, etc. In those circumstances, an average listener is not going to notice a little swishiness in the cymbals, or lack of crispness on the trumpet's timbre.
Those who care (like me) will shell out the extra bucks for higher fidelity. Those who don't care, which are in the majority, will use whatever technology is most convenient.
My guitar chord generator.
This reminds me of the fuss that currently exists over HD-TV. People gasp at the quality of the picture, but don't notice the lack-of-quality of the content. It's the same with music - people focus too much on the equipment, and ignore the music.
I've got a beautiful violin recording from the 20s or 30s. It's very low-fi, scratchy as hell, but the playing is magnificent. Ask any jazz fan whether they'd prefer to listen to a well-used John Coltrane LP, or Kenny G in 192 kHz / 24-bit, DVD-A.
People, listen to the music -- not the equipment! Otherwise you're a hifi-collector, not a music fan.
Have you heard any recent CD?!
I'd say that 90% of all new CDs have less than 6dB of dynamic range... and clip at every crescendo. I think they're mastered by people whose previous careers had them working with jackhammers without protection.
We can record in 24/192 all we want, but compression of the final product is rather moot when most of the damage was inflicted during mastering... where the "engineers" make the song as loud as humanly possible, so it could be used to silence thoughts while blasting 100dB through $5 earbuds.
It's not the engineers that are doing the mastering that are causing recent albums to be run that loud. It's the record folks that don't want their music to sound "quieter" than the competition. The engineers are just as pissed about it, but if they don't do it, someone else will, and they need to work, so they do it.
"You know your god is man-made when he hates all the same people you do."