both kinds of compression on npr

via www.npr.org

You can count on NPR to drop a cool bomb on the last day of the decade. In a long piece called "The Loudness Wars: Why Music Sounds Worse," Robert Siegel interviews mastering engineer Bob Ludwig about AUDIO compression, which is crushing dynamics out of recorded music by making it louder. Then he talks to Dr. Andrew Oxenham about DIGITAL compression, which makes computer music files smaller by throwing out data deemed to make "inaudible" sound.

Until recently, these related issues were treated as the exotic terrain of audio geeks, but people are catching on that it's a rich story about nothing less than how we use - and abuse - our ears. 

Unfortunately, I was troubled by Dr. Oxenham's answer at the very end of the story that we don't lose anything when we shrink wrap full-resolution digital music as an mp3. He says:

"Under proper listening conditions — if it's really indistinguishable from the CD as far as your ear is concerned — then you really haven't lost anything perceptually."

I can't square this with anything I've learned about perception. 

Music contains abundant information that we don't "hear" but that is perceived and used by the brain. For example, you'd be hard pressed to consciously detect the delay between sounds arriving a few nanoseconds apart, yet your brain's ability to detect that miniscule delay is what lets you know where a sound is coming from. And don't get me started on overtones and harmonics that underlie the authenticity of musical sounds and create they very basis of timbre. Those critical cues are what's thrown overboard to squash a song onto a portable hard drive and while it's fine for working out, it is exactly what leaves us unsatisfied when listening under "proper" conditions.

In Dr. Oxenham's analysis, it would be just fine to hang copies of great paintings in an art museum if the viewer couldn't tell the difference. He should want us to have access to authentic recordings so we can cultivate our hearing and become more discerning. Instead he sells short the very miraculous human sense that he studies. Peculiar. 

Comments

[Juli]

And what is lost listening to ANY recorded music versus experiencing it live?

(Isn't this also why "proper" conditions includes having a cocktail?)

[Archie Warnock]

At a certain level, what Oxenham says is true - high-resolution MP3s do sound indistinguishable from the originals to most people. The problem is that it's hard to say where the dividing line between good sound/bad sound occurs. Generally, though, most MP3s these days (with 128kb sampling) are over-compressed, trading off too much sound quality for disk space. It's not nearly as desirable now, especially considering that disk storage is extremely inexpensive.

In short, you can rip high-quality MP3s, but generally not by default. You can use lossless compression, like FLAC, and trade off a little more disk space for real high quality. You can buy high-quality MP3s from the Amazon download store.

Lots of options - it's not true that compressed audio files have to sound bad.

[George]

A lot of this discussion really has to do with the listening environment. Listening to music in a car is different from listening to the exact same recording in a quiet room or on headphones. I once sold audio equipment, and I can tell you the sound will change as you vary the monitoring gear. Since most people listen on crappy equipment in noisy surroundings, then compression is invaluable. The more, the merrier. Compare the acoustics of a concert hall with a home living room. By the same token, recording techniques today have eliminated the room from the process. All the instruments are close-mic'd, and everything is remixed and processed in the mixdown session, after the musicians leave. The end result is an unnatural sound that has become the new normal for most music listeners.