What is sound?

What is sound? Yes, it’s noise—but scientifically it’s changes in air pressure caused by vibration. The pressures it creates are inherently analog, that is, they’re defined in relation to something else and are continuous, with many possible states.

For humans, audible sounds are all within the range of 20 Hz to 20 kHz; the higher the value, the more “high-pitched” the sound. Think of low, rumbly bass frequencies as being near the lower bound of that range, and sharp feedback from a microphone being nearer the higher bound.

Analog vs. digital

Analog is the opposite of digital. If something’s digital, it has only two possible basic states: 0 (off) and 1 (on). Digital media is discrete (not continuous) and doesn’t allow for continuous values in the way analog does.

At this point, maybe you’re asking If sound is inherently analog, what does it mean for something to have digital sound? How can all the possible pressures of sound be represented by 0s and 1s?

The “digital” part refers only to how the sounds are represented on their storage medium, whether that’s a streaming service such as Apple Music or Spotify (either of which plays remotely stored files), a CD, or a Super Audio CD¹. To create digital sound, the original analog signal is sampled: Its continuous signal is reduced to a discrete one, with each sample a “snapshot” of the signal’s amplitude at a point in time. Together, these samples, which are represented as numbers and spaced evenly in time, can be reconstructed into the original signal by a digital-to-analog converter, a small computer that converts these numbers into corresponding pressures.

I pour an inch of iced tea to match the level of the rain gauge I’m eyeballing from memory. The sound tumbles over the cubes below, its noisy current apparently familiar with the thunder, as if it’d sampled it at a low rate and shallow bit depth. Back in the waiting room before college, a lake that’d once swirled with the entirety of adulthood in its matching-color eye and beckoned for me to drown within it, had been Aral Sea’d, exposed as the aquatic emperor with its bed cracks visible to all. “Curvature, the earth, unevenness, and the wind”—a teacher’s words came through discretely as I digitally tested the lakebed for a hole to Wonderland and imagined it as the now-drunk-to-nothing cauldron of the first tea party. A Neanderthal atop a mastodon and a giant rabbit once used elaborate sluice gates, worked by pre-steam engine empowered labor, to fill this pond with glacier water and chunks and the hibiscus flowers whose pedals bled into something that already transcended Southern-ness.

If I wanted to accurately capture the sound of clinking ice cubes, let alone a collection of instruments playing simultaneously, on a digital medium, then two things determine the fidelity of my recording:

• The bit depth: You’ve likely heard something described as “8-bit,” “16-bit,” or “64-bit.” Without going too much into mathematics, this descriptor, called the bit depth, tells you how many possible values there are in the computing system in question. For the purposes of digital sound, it denotes how many possible discrete steps there are for each sample to be assigned to. For example, 16-bit audio like the standard CD has 65,536 possible steps that a sample can be assigned to.
• The sampling rate: This is simply how frequently I’m sampling (taking aural snapshots of) the original sine wave. For a CD, the sampling rate is 44,100 Hz—a strange-seeming number that nevertheless has some logic behind it: It’s the product of the squares of the first four prime numbers (2, 3, 5, and 7), which makes prime factorization easier.

Losing it over lossy-ness

What about “lossy” audio formats? What’s that, you didn’t know that most music you listen to has “lost” some of its original character, because its file size has been greatly diminished or at least compressed so that it can be transmitted over a network (like the Internet)?

MP3s, which dominated the 2000s with the advent of the iPod, are often 90% smaller than the on-disc files they were ripped from. Modern streaming services use lossy files, too, although Apple Music at least offers lossless quality—but you won’t actually get it unless you use wired headphones. Most implementations of Bluetooth are lossy and they always compress the file. Some wireless forms of transmission, like AirPlay-ing a song from an iPhone to a HomePod, are lossless because they use Wi-Fi instead of Bluetooth.

Does lossless music sound better, though? You may think there’s a huge gap in quality given the order-of-magnitude difference in size. But the human ear isn’t that discerning. The range of frequencies that a standard CD can reproduce—from 20 Hz (below the lowest key on a piano) to 20 KHz (something that most adults can’t hear)—is already outside the extremes of human perception. Plus, the algorithms that create lossy files are optimized to remove the parts we notice the least.

I’m back from the lake, soaked with the feeling of needing to look into something just as sensuous as its bygone fullness, so I’ve got a curved 1996 TV with windy signals sweeping across its face. In a later life, it’ll be dusty, birthed anew though old with nostalgic reverence, but right now it’s pristine and basic, with speakers that yield from a cartridge-loading game console a no-temperature no-color sound—anti-adjective waves, sweeping past my enlarged elf-like ears, red from excitement, with the spirit of being and doing rather than of looking and sounding and feeling. My ears once pulsing with analog cinders are cooled by digital rain that angles its math earthward. The sound began in the CO2-exhaling fires of circuit board production and then a factory blew it across the sea on a blimp of emissions, harmless by all appearances in its big boxy cartridge with just smokeless ones and zeroes inside, etched errorlessly on metal. Next up there’s a truck trek from the dock and some plastic to unwrap and discard to live an eternal life unseen, shining curvily and brokenly like that TV the moment some hypermodern post-man Nouveau Neanderthal digs it up despite the newly Venus-like pressure on Earth.

Though the struggle is seemingly won by lossy formats, what if on some other timeline the exact opposite had prevailed? What about a format that was somehow even “realer” than CD? Enter the…

Super Audio CD

Super Audio CD was meant to succeed the CD as the dominant physical music format. Most people don’t know what it is, and that’s understandable—it arrived right as the music industry was about to collapse into piracy and financialization and it confronted you with two things you’d expect from such a harbinger: Few noticeable improvements for its high cost, and aggressive copyright controls.

SACD was meant to kickstart another cycle of format replacements. The CD was approaching its 20th anniversary, cassette tapes were dying, and vinyl was dead enough that had yet to be zombified back into existence. The maturity of the CD lifecycle, as Andrew Dewaard describes in his book Derivative Media, arrived in the 1990s when there was already a nascent trend of trying to financialize music by making it more like rent (I’m sorry, a “service,” which all digital things seem to be becoming²), something that pointed to the demise of one-time physical disc and tape purchases:

Transforming music royalties into an investment strategy is not a new idea; David Bowie even sold “Bowie Bonds” to investors in 1997, based on income generated from his back catalog. “For the music industry the age of manufacture is now over,” Simon Frith claimed back in 1988, as music companies were “no longer organized around making things but depend on the creation of rights.” What is new is that those rights are now much more lucrative and have attracted much bigger financiers. As opposed to physical media, which was typically purchased only once per format, listening to music on a streaming service produces a financial transaction every time a song is played, dramatically increasing the value of older music.

The surge in piracy that accompanied the popularization of Napster in 1999-2000 undoubtedly hurt the music industry, but only one segment of it: artists. The profits they could make from selling immensely profitable LPs, CDs, and cassettes evaporated. Meanwhile, laws such as the Digital Millennium Copyright Act plus DRM controls in iTunes and the eventual “you have no control over this” total platform control of streaming services such as Spotify and Apple Music empowered labels, who exploited this crisis and funneled the entire music industry through their subscription services while paying out literal fractions of a penny for each stream.³

This labor-hostile reality, in which musicians are barely compensated for their music, skews listening to musicians who are already famous and favored by the highly centralized software algorithms that underpin streaming service interfaces. Breaking through thanks to regional radio or an unexpectedly popular CD release is no longer an option.

Given today’s context and how it compares to what came before, SACD looks now like a strange transitional form between modes of capitalist exploitation: The ownable physical product, of which it was the real attempt to create a mass-market optical disc format for audio⁴, and the locked-down streaming future:

• Unlike CDs, which were unencrypted, SACDs were and still are a pain to rip. They’re not recognizable by standard Windows or macOS optical drives, let alone standard CD players. You need dedicated playback hardware or else a pricey industrial solution, just as you need to pay money forever to maintain access to a music streaming service.
• But all this protection was meant in part to prevent the type of broad (and free) distribution that CDs had enabled. You couldn’t make CD-R copies of SACD discs and couldn’t easily leak their tracks to the internet.
• They also aimed to go even further than CDs had on the quality front, by radically changing both the bit depth and sampling rate components mentioned above. SACD lowers the bit depth to just 1-bit (so everything is literally “on” or “off”), but bumps the sampling rate up to over 2 MHz. So there are many, many more samples of the analog signal, but instead of each sample storing a step that could fall along thousands of possible values, it just indicates whether it’s higher or lower than the one before it. This setup was branded as Direct Stream Digital (DSD) and was designed to give a more analog feel to digital recordings.
• Finally, it allowed six discrete channels instead of the two (“stereo”) of CD. SACDs can output up to 5.1, meaning there’s the two stereo channels plus a center channel (usually for vocals), two rear channels, and a subwoofer.

The in-betweenness of SACD is embodied by how it’s a digital format that for years was easiest to play over analog connectors such as RCA jacks. The very first SACD player, the Sony SCD-1, could only play SACDs over either RCA or XLR connectors, both of which are analog. The digital optical and coaxial connectors couldn’t be used with SACD discs because they lacked bandwidth—as well as copyright protection.⁵ Modern hardware that can play SACDs, such as some Blu-ray players, can usually only do so via HDMI, which requires a receiver that can decode DSD, which can’t be assumed.

And for what? Does it sound better? SACD was a pivotal moment in the software-ization of music, in that its new features—aside from the 5.1 sound, if you had the right setup for it—had the inscrutability and utilitarianism of a software version update or patch. It was “moving forward” for its own sake, browbeating and exhausting listeners who had long since stopped noticing the belated improvements. Except here, they had the choice not to endure it because of the newer and cheaper MP3 and streaming markets; there’s something optimistic there, in that it reminds us that maybe someday we could break free from how shitty software has made our entire world.

When I first saw you, I knew you possessed the fiery shape to imprint yourself on my memory as if I were an inanimate disk drive—a metal machine marcher who remembers things even while unalive. Someday I too will be dust blown over a landfill, brought to faraway dumped-out ears in perfect surround sound with enough ambient heat to case an earbleed. There’s a new sun directly on earth; we’re moving through time, backward to a pre-Neanderthal era that never knew agriculture, just a forest where a woman is running in the dark underneath a sheet of rain.