For the best sound quality, does the format really matter?
The Audiophiliac ponders the ins and outs of high-resolution audio.
High-resolution audio is frequently touted as the best possible way to experience music. Maybe, but I also think that standard resolution CDs and 320 MP3s, ALAC, and FLAC files can sound good, really good. Before we dive into the numbers I want to make the point that sound quality is first determined by the quality of the gear used to record, mix and master music, and perhaps even more important, the skill and artistry of the engineers. Those factors influence sound quality far more than the release format. In other words, a great recording on MP3 will definitely sound better than an overly compressed and processed one as a 192 kHz sample rate, 24 bit FLAC. If the original recording quality was crap, it's always going to sound like crap. Sure, it would be so much easier if we could predict a recording's quality with a number, but the numbers just define the limits of how good it could be, not what it will be.
Grappling with the concept of high resolution audio might be too abstract, so it might make sense to compare it to high definition TVs. They produce sharper, more vivid images than standard resolution TVs, but a blurry old movie, transferred to a 1080P Blu-ray, won't make the film any clearer. So if the film is murky, a high definition TV won't make it any sharper.
A high-resolution audio file's "sample rate" refers to the number of samples recorded per second when the analog sound waves were converted into a digital file. A 192kHz sample rate recording has twice as many samples per second as a 96kHz recording, which is more than double the sample rate of a CD. A 24-bit recording has more bits than a 16-bit recording, duh, and those extra samples and bits can more accurately capture the subtle stuff, like the breath of a vocalist or the ambient sound of a recording venue. While the advantage of more bits is nice in theory the benefits may be lost when the recording, mix, or mastering engineers crush the instruments' and vocalists' soft-to-loud dynamics. More samples and/or bits won't restore the missing information.
Why are recordings handicapped this way? I asked my friend, mastering engineer Alan Silverman about it, and he described a "feedback loop" where producers, artists and engineers make their way through the studio process to release an album, and how they later hear it in the real world influences the way they'll make their next album. That's the way it worked in the 1950s, with LPs and singles, which were heard at home on a hi-fi and on the radio, so the engineers made recordings that sounded best at home and on the radio. That initially held true with CDs, but now that most people experience music away from home, in noisy environments, on lossy streaming services like Spotify, all of that informs the way producers, artists, and engineers approach their next recording. Silverman believes that the way today's pop music is arranged, designed and performed is strongly influenced by the quest for extreme loudness and limited dynamic range required for music listened to on-the-go with tablets, ear buds and phones. It's one thing to hear the sound in the studio, but the music will mostly be "consumed" as background filler. It's rarely the prime focus anymore.
Silverman also said, "I'm not sure anyone has the definitive answer as to how, or even if, high resolution audio sounds better. The answer lies deep in digital signal theory, psychoacoustics, and signal detection theory, which is the science of testing subjective responses. BTW, the latter, which is critical to answering the question, is entirely ignored by those running blind A/B tests to bolster their claim that high resolution audio does not matter. I can only speak from my own experience working with audio and music. High-resolution audio feels better to me. After working in hi-res on a project for an extended time, and then down-converting the project to standard resolution to prepare the record for release, I can feel the difference, dramatically at first, without even thinking about it. On that first playback at lower resolution something has changed and then you remind yourself, oh right, it's down-converted now. Then as I get accustomed to the lower resolution, the difference seems to recede, but I am certain I heard it at first. Since part of my job as a mastering engineer is to find what I can do with the music to make it feel better and enjoy it more, I try to find a way to translate that to the end-listener. I'm trying to maintain the best qualities of high-resolution sound from production to mastering to release, so it makes sense in my world."
High-resolution files are more expensive than standard definition files or CDs, and they're available from a number of websites. So the question is raised, is high resolution worth paying extra for? That depends on a number of factors, first do you have a really decent set of headphones or speakers? The benefits of high resolution audio can be pretty subtle, and you have to really listen to appreciate the extra detail. Next, is the recording good enough to benefit from high resolution mastering? Some are, most aren't, but when it's done right high resolution audio presents a clearer "window" to the music, it's a real thrill.
If you'd like to listen primarily on a phone, you'll need a phone that can play high resolution files, like the new HTC One M8 Harman Kardon Edition. Better yet, consider a high resolution music player, such as the Astell & Kern AK100, Hifiman HM-901, FiiO X3 or X5, or the PonoPlayer that's due in October. You can of course play high-resolution files on most computers, but to get the best sound I recommend purchasing an external digital converter, like the $350 Schiit Audio Bifrost. Sony recently got on board with a new push for high resolution audio and introduced two home players.
I'd love to hear from readers currently listening to high-resolution music, tell us all about your experiences in the comments section.
You can subscribe to the Audiophiliac RSS feed here.
Update, 11:06 a.m. PT: Adds quoted material from Silverman, for clarification.