I’ve lamented in reviews over and over and over again about compression in modern recording. Often readers misunderstand what I mean by compression. Consider this an effort to try to clear the air a bit and help you understand why I’m complaining (well, why I’m complaining about this, anyway).
First of all, there is audio data compression (also known as file compression). This is the process of “compressing” the digital information found in a computer file into a smaller size. To accomplish this with an audio file, specific data is removed from the file structure. Picture a tower of Legos that stands 8 feet tall. You need to fit it into a 6-foot-high space without destroying the tower, so you strategically remove Lego blocks here and there to scale it down. That’s what data compression does – it strategically removes bits and pieces from an audio file to make the overall size smaller. That’s why mp3’s are significantly smaller than wav files – they have fewer pieces of data in them.
That’s also why mp3’s can sometimes carry a sub-par sound. For the most part, the digital pieces removed from a “lossless” wav file are often undetectable to the standard listener, especially if the mp3 is at a higher bitrate. If you start dropping below 256 Kbps, however, you can start to notice a reduction in overall quality. The lower the bitrate, the lower the quality (because you’re removing more pieces of the file).
So why don’t songs downloaded from iTunes or streamed from Spotify sound different compared to a CD? Well, there’s two reasons for that. First of all, when you’re listening on a pair of plastic ear buds or through a smartphone speaker, you’re not going to notice the little audio details that may or may not be there (part of the reason Apple isn’t in a huge hurry to update their earbuds to a higher listening quality). Secondly, most digital music vendors have a minimum quality requirement for their music files. If you buy an mp3 from Amazon, odds are it’s going to be a minimum of 256 Kbps (if not 320). In fact, when an artist submits an album to a digital distributor such as CDBaby, they actually require that you upload 1411 Kbps wav files (the standard CD-quality bitrate), and then the retailers can downscale them according to their own standards.
Dynamic range compression (which is what I usually refer to), on the other hand, is something different, and has to do with volume levels, not file size, although both involve making changes to an audio file.
I’ve shared this video before, but in a nutshell, here’s how dynamic range compression works:
As you can see, this type of compression is the practice of taking volume changes (dynamics) and “compressing” them into a narrower volume range. Instead of having songs with loud parts and quiet parts, you have everything at one set volume level, which not only takes away from the musical dynamics, but frequently leads to distortion, as the louder volumes (such as the drum hits in the above video) are being limited.
For comparison, imagine standing in an open, empty gym. If you start singing, the sound will bounce all around and create lots of different effects for the listener. Now imagine standing in a closed hallway with fabric on the walls and carpet on the floor. If you start singing in there, the echo effect of the gym is gone because the sound has been “compressed” into a smaller space.
That doesn’t mean all audio compression is bad. When used properly, it can help with flaws in the natural sound of both instruments and vocals to help balance out an otherwise good performance. However, modern recordings have taken this concept to the extreme by compressing entire songs, creating (as the video above called it) “wimpy loud sound.”
Now, surely record producers and engineers know what this is doing to recording quality, right? The answer is yes, of course they do. So why is this done so often? Who is the blame?
The simple answer is, we are. The listeners. As a whole, music consumers do not “actively” listen to music anymore.
Before the days of portable music players, if one bought a recording, listening to it was an experience. Stereos were often significant pieces of furniture that had their place in the living room, or perhaps a smaller record player or tape player was found in your bedroom. When you bought a record (or a tape or even CD), opened the packaging, put it into a player, and you sat and listened. It was an activity.
When Sony introduced the Walkman portable tape player, all that changed. Listening to music evolved from being a primary activity to “something in the background.” Although the formats have changed, the same basic concept is there – load your favorite songs on a tape, a CD, or a digital playlist, and you can play the songs while you work, exercise, or any other activity. If a song had too many dynamics in it, it was considered a distraction to whatever the listener was doing, so the volume had to be edited in such a way that the sound doesn’t fluctuate. Enter audio compression.
Nowadays, we listen to music in ear buds of varying quality while we do everything else. We don’t notice that a recording may sound subpar; we only notice when it’s too soft. In fact, as consumers, we’ve become so accustomed to how audio compression sounds on recordings that, if it were to be removed or cut back, we would think something was wrong. When I reviewed a recent compilation of songs from the 1970’s, I had one person ask me later if I thought the album was “a little quiet.” I said, “Yes, it is, because it’s not being filtered through modern-day audio compression and made as loud as possible.” They knew something was different, but they couldn’t quite figure out why.
And yes, this happens in gospel music as well. For example, listen to this recording of “It’s Almost Over” by the Cathedrals from 1984:
https://www.youtube.com/watch?v=bqu9Q-lpBIk
The intro (as played by the late Lari Goss), is almost undetectable, and that’s on purpose – it’s supposed to build from virtually nothing to a large, epic ending. Compare this to the Mark Trammell Quartet recording made in 2010:
This intro (mostly unchanged) is far louder in this recording. Some may say the dynamics are found in the lack of other instruments, but the overall volume is fairly consistent.
As it stands now, some gospel groups still take it to the extreme, creating an overly-artificial sound (for lack of a better term, it sounds “plastic” to me). It is my hope one day that audio compression (much like auto-tune and digital phrasing perfection) will one day be reigned in a bit and brought under control, allowing for natural talent and abilities (even if slightly imperfect) to one day shine through.
Great explanation and example. Now I know why I don’t buy much music any more and enjoy my old recordings more.
The compilation from the 70s was quite refreshing, and I’m glad the guy who mixed it stayed true to the era.
KB: “it is {also} my hope one day that audio compression (much like auto-tune and digital phrasing perfection) will one day be reigned in a bit {a lot} and brought under control, allowing for natural talent and abilities (even if slightly imperfect) to one day shine.” AMEN!
I very much appreciate these days when I (rarely) hear a quartet whose songs have not been electronically modified. It would be great to hear what these groups sound like live and in-person–but on a DVD.
Interesting and well written article. I don’t detect the subtle differences on my own but I can tell a difference when you point out what it is. I have several hundred CDs, maybe 100 cassettes, and even some vinyl records. I seldom listen to any of them because I buy most of my music online now and download it because it’s so convenient. I may not have the quality of sound but I can play any of several hundred songs any time because I always have them with me. I’m always looking for new music that’s done well.
I see this all the time in the studio, especially when compiling songs for a radio compilation and pulling the songs from 10-12 different artists and projects. The WAV files all vary … some look like a “2 x 4” and others were left with “breathing room” and look like a pretty mountain range…all depends on who mastered that particular record. Personally, I like seeing the dynamic range left in a recording … good article and good explanation!