On Audio Quality / by Sam McArthur

Most people older than 20 will remember a time when middle-class America listened to most of their music on cassette tapes or over FM radio. Compact discs had been around since the early '80s, sure, but even as recent as the mid-1990s, buying an album on CD instead of cassette was a luxury to many average people. To those same people, there was very little transitional time between cassette and MP3 around the turn of the century. It was all down to price, and even then, a cassette was still quite a bit less cash than an album on MP3. Now? 25 cent cassettes at Bull Moose, baby.

Back in the dark ages of the 1970s, Most music purchases were a big deal. Audio cassette hadn't yet seen its spike in popularity following the downfall of the 8-track, so much of the home audio market was vinyl, which was expensive, but at least it sounded good. On the recording side, studios often used microphones designed in the 1960s, many powered with vacuum tubes, going into shiny new recording consoles like a Neve 8038 or an API 3288, and then onto a tape machine; either 16-channel, or 24-channel if you were so fortunate. Compared to the technology of just ten years prior, the final products sounded stunning, between the widespread introduction of stereo releases, greatly improved frequency response, and a lower noise floor... and that's to say nothing of the decade's raw songwriting talent. 

Most of those songs are now more than forty years old. They still sound fantastic, compared even to our most modern music. 

Why is that?

Well to start, it's not due to the technology. Even middling digital recording technology records audio with such a pristine capture of the electrical impulses given to it, and with such a low noise floor, that analog tape recording is technological prehistory. The conversion back into an analog signal is a much trickier game, because even small variances in the timescale between samples starts to dramatically affect our perception of the sound quality. Most companies put a stronger focus on the Analog to Digital conversion because it's the most important part of the digital signal chain: In the computer, your music is pure math, each computer treating it precisely the same, regardless of how it leaves the computer and reaches your ears. If you mix completely within Pro Tools, Logic, or any other DAW, the sound should still be mathematically perfect.

So again, why does music from the 1970s sound so starkly better than most modern music? In short, it's mostly down to what happens after that song leaves for distribution.

The best-case scenario that most albums can ask for is to get a dedicated mastering for their CD release. This will normally leave a bit more dynamic range on the songs than if you listen to the same song on the internet. A good CD release will reproduce the original track objectively better than any MP3 download, or even a record. The reason a record is so aesthetically pleasing is because of the nonlinearity introduced by the physical grooves and the electroacoustic response of an actual piece of metal reacting to them. Our ears and our brains like this nonlinearity... That's the "mostly down to" concession. 

That said, most people listen to music downloaded from Amazon Music, iTunes, or streamed from one of any number of websites. To save on hard drive space and download time, each song is put through mathematical processes that takes into account which parts of the sound you'll hear, and which your brain can put back together, and rip out everything that it doesn't absolutely need. Depending on where you're listening, the quality will be more or less tolerable. In this realm, sound quality hinges on the amount of kilobytes that the service will push to you per second. YouTube, played back in HD, tends to have above-average quality. While the actual kbps that YouTube uses is obfuscated and hidden away, it's definitely somewhere below CD quality, but still comfortably listenable on an iPhone or your laptop's speakers. iTunes' current format sounds about equivalent, and runs at 256 kbps. On the opposite end of this spectrum, Pandora's streaming service provides 192 kbps to its "Pandora One" subscribers, and a whopping 64 kbps to free users.

In addition to that, most masters for digital format nowadays tends to slam the songs through a few compressors to wring the most volume possible out of the poor ones and zeros. This may seem to make the music more energetic on first listen, but compared to older masters, it's missing a lot of punch and dynamics that make music sound good. Find a song from the early '80s on its original CD release and that same song from a recent remaster, and play them at the same average volume. The newer track will start to feel physically draining, while the older version continues to be enjoyable. This is just as true for any highly-compressed modern release as it is for music that was released decades ago and remastered: If it's loud, it's going to get exhausting. 

Now that you know the basics of why so much modern music sounds like garbage, what can you do about it?

If you're a listener who's starting to care about fidelity, buy on CD. If you can, buy a vinyl release. A compact disc is the best-sounding mainstream format for the money, and it's made even better thanks to the high-quality Digital-to-Analog converters in most CD players. On vinyl, you can be basically assured that the master can't have been compressed within an inch of its life, due to the limitations of cutting sound into a physical medium and playing it back with a needle.

If you're an artist and you're going into the studio, start by insisting on recording at a high sample rate. At The Greenhouse, we do as much recording at 96 KHz, 24-Bit as possible. The difference between CD quality (44.1KHz) and 96KHz, is much more dramatic than the difference between 96 and 192, and with 192 you'd be using twice as much storage as 96. However, if storage or read/write speed is a concern, 48k at 24-Bit Depth is a more than acceptable quality for most releases.

After having your great music recorded with high-quality microphones and front-end gear into a quality audio interface at a high sample rate, you have two options on how to proceed. First, you could keep your mixing in the digital domain to retain that digital perfection. Second, you could do as much of your mixing in analog to reintroduce the idiosyncrasies and imperfections inherent in analog hardware that makes classic albums so pleasing to listen to. Of course, then the converters will again become much more important. We use a pair of Antelope Orion 32+ interfaces, which aside from modified units and the truly state-of-the-art, have some of the most pristine and true-to-source converters in the business, both in and out. 

Once your music is recorded and mixed, you'll be taking it to get mastered. Start by getting the final mixes in the same sample rate they were recorded at. Going lower will degrade quality, and going higher is just a waste of space; you won't add any information during mastering. Find an experienced mastering engineer who won't go straight for a super-loud "Consumption-Ready" product, and will retain the sample rate and quality of your songs. This is easy as an independent artist, but if you're signed to a label who tries to handle all of this distribution-related stuff themselves, you may have to be insistent that you get someone who respects the audio fidelity, even if you aren't planning on a vinyl release. Keep in mind that while there will always be a balance point between commercial-friendly loudness and dynamics, where that exact point is depends on the style of music and how many elements (instruments) are within your song. Sure, your music may be quieter overall than other albums, but again, once people crank the volume up to comparable average levels, your songs will have the life that other much louder releases lack. 

At the end of the day, artists that care about audio fidelity make albums that please listeners that want audio fidelity. It doesn't matter too much what the experience of an "average" pop music listener will be, because their listening conditions are often the lowest common denominator and their most common services (Spotify, FM or Satellite radio) are already compressed to high heaven, both with acoustic compression and format compression. If you have well-written songs that have been recorded, mixed, and mastered by people who care about quality and know their stuff, then you'll have an end product that has the chance to stand the weathering test of time like your favorite albums from the Analog Era. 

Recording to tape? That's a different story.