The perfect-sound myth

If digital recordings were truly better than analog, they would sound more like real music, but digital just sounds different, not better.

Photo by Steve Guttenberg/CNET

I remember just before the CD was introduced 30 years ago thinking that digital audio would be a giant leap forward in fidelity, but as soon as I heard a few CDs I knew digital wouldn't do a thing to make music sound more realistic. The CD was vastly better than LPs and cassettes in terms of noise and distortion, but voices still didn't sound like they do in real life, and pianos didn't sound as big and powerful as they do in Carnegie Hall. That mystified me; those early digital recordings were compression-free, and I was told digital didn't add or subtract anything from the sound the microphones recorded. Digital sound should have been perfect, but it was just different than the analog recordings I grew up with.

That's when I realized that striving for ever more accurate recordings wouldn't improve sound quality. The things that make sound pleasing to the ear aren't limited to making technically better recordings (or hi-fis). Great-sounding recordings sound great mostly because of the hundreds or thousands of decisions made by the engineers who recorded, mixed, and mastered the music. Their choice of using a microphone that flattered the vocalist or saxophone, the acoustics of the recording venue, the processing that was used to create each sound within the mix make or break the sound. The recording format also plays a role, but analog or digital, they're just a small part of the overall sound picture. Perfect sound isn't really what most engineers are striving for; they just want to make a recording that sounds good. And good sound is a purely subjective call.

High-resolution 96-kHz/24-bit or 192-kHz/24-bit digital recordings only produce incremental improvements in the believability of sound over CDs; the other factors mentioned above play much larger roles in sound quality than higher-resolution audio. Neil Young's crusade to introduce a new high-resolution format is well intentioned, but misses the fundamental problem of how the original recordings sounded in the first place. Play a heavily compressed and equalized recording in Young's Pono player and it's still going to sound like crap. The old "garbage in, garbage out" truism definitely applies. I'd rather listen to a great-sounding recording as an MP3 than an awful-sounding one in a high-resolution player. Technical perfection is one thing; making lifelike recordings can't be reduced to a numbers game. Great sound is more of an artistic than technical pursuit. Sadly, Young's device won't do a thing to fix the recording's inherent flaws. It has to sound good to start with.

The same criteria apply to speakers and amplifiers; if designing great audio gear was just about reducing distortion, we would already have reached the goal of reproducing perfect sound. I've heard a lot of the world's best audio gear, and even there, nothing comes close to reproducing the sound of an orchestra or a rock band. I can't imagine a major design breakthrough over the next decade or two that will change that, the recordings are the main obstacle. If the sound isn't fully captured in the first place, a perfect set of speakers and electronics won't be able to reproduce the true sounds of voices and instruments. As it stands now, analog or digital recording technologies aren't the limiting factors.

I covered this subject in greater depth in a recent issue of Stereophile magazine.