CD vs. mp3 ... a professional test

The reviews mean nothing, like the prizes. And if it's not well installed, it can't work as well.

Edited by oliverstoned - June 01 2007 at 08:47

c't The Magazine for Computer Technique  June/2000, p. 92: MP3-Comparison  By Carsten Meyer  Original German article: http://www.heise.de/ct/00/06/092/   Cross-examination test  The c't-Reader’s Listening Test: MP3 versus CD

Hello sean! Nice to hear from you again.

And let me post that article:

MP3 & the Marginalization of High End Audio

John Atkinson, February, 1999

It was the weirdest orchestral balance I'd ever heard. The gentle woodwind chords that begin Mendelssohn's incidental music to A Midsummer Night's Dream were as loud as the climactic "Wedding March" that ends the piece. The radio broadcast was obviously being compressed to hell. Yet, sitting at the wheel of the rented Vauxhall Vectra I was driving down to Cornwall for an old friend's surprise 50th birthday party, I was actually glad for the compression. Had Classic FM broadcast the Mendelssohn with its true dynamic range intact, the quiet passages would have been irretrievably buried in the road noise and the loud passages would have had me lunging for the volume control, to the possible danger of those sharing England's congested A303 trunk road with me.

Compression has its place when mandated by suboptimal playback circumstances, such as in a car. But as compression also raises the average energy levels of recordings, thus making them sound louder without their peak levels being changed, its use is ubiquitous when CDs are mastered to make them sound more aggressive on the radio.

Take this issue's "Recording of the Month," for example: Mutations, from post-modern waif Beck. Of the contenders music editor Robert Baird had gathered for this issue's feature, I voted for this CD on musical grounds. And yes, Mutations does sound very clean, and the mixes have been done with musically literate intelligence. But, as revealed by the professional Dorrough loudness meter I use for my own mastering and mixing, the recording's dynamic range (footnote 1) is only a few decibels. Compare that with Stereophile's Duet CD, which has a dynamic range of well over 40dB—a 100:1 level ratio.

I am fortunate to have grown up in a family involved in live music making, and to have spent my formative time playing in orchestras and other acoustic ensembles. And my active involvement in rock music came at a time when the sound of acoustic, unmiked drums set the volume level for the rest of the band. I know what the real thing sounds like, and I can recognize the effects of studio tricks like compression. (Although, as you'll read in the article on the recording of Stereophile's new jazz CD, Rendezvous, in the March issue, I am not averse to their judicious use.) However, if all people listen to at the turn of the millennium is rock recordings like the new Beck album, and Classic FM and its American equivalents, then I begin to worry about the purpose of the High End.

What, for example, is the point of spending a large sum of money on a loudspeaker like the Revel Gem, which I reviewed last October? The Gem was engineered to minimize compression, but if the music its owner likes to play has had the dynamic range engineered out of it, the advance in speaker performance represented by the Gem is rendered moot.

And if listeners are no longer in touch with what the real thing sounds like, can it be surprising that a medium that actively abandons quality can become successful? In his item on Internet audio in "Industry Update" (p.35), Jon Iverson states that "MP3-formatted audio files are considered to be the most popular streaming technology on the Internet." It could be argued that MP3 has been the most significant introduction of a new audio medium since the CD 16 years ago. Yet not only has Stereophile almost completely ignored the MP3 audio format, so has every other audio magazine—as has the high-end audio industry itself.

But of course, you murmur, MP3 (MPEG 1 or 2, Layer 3) encoding is such an aggressive data-reduction algorithm that it does not produce sonic results that could be defined as high fidelity. Why, it even treats the basic signal as mono, the stereo illusion being provided by an attempt to encode the difference between the two stereo channels. If it isn't high fidelity, then an industry that strives for the highest fidelity shouldn't bother with it.

But when we turn our backs on what our potential customers and readers are doing, don't we marginalize ourselves—particularly as we audiophiles tend to concentrate on the tools, not the experience? As Kinhluan Nguyenngoc posted on The Audiophile Network back in March 1994, "I believe the rags are leading music lovers down the wrong, component-centric path—buy the 'right' component, and everything will be copacetic."

As if that weren't enough for high-end audio, societal evolution seems to be marginalizing the act of listening to music for its own sake in favor of music as a background activity. If my own experience is anything to go by, the increasing bombardment of information—from traditional media as well as from the Internet—makes it harder to budget the time to listen seriously to one, two, or three hours of music at a time. Yet I do manage to budget that time. But when I talk to multitasking Generation Xers or Yers, they look askance at me for devoting so much time to a single activity. "They think I'm nuts for just sitting and listening," an acquaintance agreed when I mentioned this to him.

Perhaps there is something antisocial about cocooning oneself in a room dedicated to recorded music with a solitary chair in the sweet spot, compared with sitting with a bunch of friends watching a new movie on DVD with good, enveloping surround sound. Again, it looks as though we audiophiles are marginalizing ourselves.

So if quality in sound reproduction is out of fashion and solitary listening to music as a goal is no longer normal or even socially acceptable behavior, then is it any surprise that in the midst of a boom in consumer spending, sales of two-channel audio equipment are flat? Is there any reason to think that high-end audio reproduction has a future at all?

Yes. High-end audio is not dead, or even dying. As I reported last month, the introduction of DVD-Audio, with its combination of inherently better-than-CD quality combined with a native multichannel format, will re-excite the public's interest in audio quality. Sales may currently be flat, but music is fundamentally important to human existence; as an essential adjunct to music, high-end audio will always be with us. In fact, although more people are listening to poor-quality sound than ever before—thanks to MP3, cynically designed boomboxes, and shoddy all-in-one systems—the average quality of audio is higher than it has ever been, and the market for high-quality audio equipment is bigger than it has ever been.

As I've pointed out in recent "As We See It" essays, it's not that the potential customers for high-end audio components have disappeared, but that the high-end audio industry has lost sight of ways to reach those customers.


Footnote 1: Dynamic range is defined here as the difference between the peak levels of the loud and soft passages. Dynamic range is more formally defined as the difference between the peak level and the noise floor.

To conclude: digital was a regression, sub-digital is even worst...

I've done better: to compare a burned-computer Cd versus the orignal, which is, according to your theories, the exact copy.

oliverstoned wrote: The reviews mean nothing, like the prizes. And if it's not well installed, it can't work as well.

Edited by Snow Dog - June 01 2007 at 13:58

This is far too interesting and no other posters should interfere - keep it up guys, you have an obligation to entertain.

Mike - don't tell me I said something you approved of?

For me the Mp3 format is excellent when used from the bands/ artists/ producers (or from Internet site) in stream version to promote songs or video. Otherwise I find more interesting the CD format.