Mediocre music and the mainstream market

The music you've come to love has been sabotaged.

Not many people realise the degree of degradation music quality has experienced over the years. It has got to the point where the old vinyl records of yesteryear, the same ones that were popular some 100 years ago, provide higher quality dynamic range and clarity than today's CD audio. Remasters of old classics are sounding worse than the originals, and new music is sub-standard from the get go. It begs the question; surely technology has improved since vinyl?

In fact, it's improved immensely. So why has quality gone backwards of late? It's because of a shift in priorities in the mainstream market. Consumers have been conditioned to enjoy highly compressed low quality audio, to the extent where the original sound is significantly different from the original. With the advent of portable digital media players, the lossy audio format knows as MP3 (MPEG-1 Audio Layer 3) has become a widespread hit. It can convert a typical CD quality lossless song (say, 30MB) into a handy 3MB file size at 128kbps, or 9MB at 320kbps (the highest bitrate that MP3 facilitates). That's ten times smaller. 30MB might not sound like a lot of data in this day and age, but when people have music collections well into the thousands, it all adds up; and therein lies the problem. People don't listen to thousand of songs on their music player, it's a matter of convenience. Why transfer songs as you need them when you can have your whole library with you wherever you are?

Flash storage in today's compact Portable Media Players (PMPs) is still quite limited compared to the vast amount of music available to us. 32GB is the upper limit of these players, and they're not exactly cheap. You can fit over one thousand lossless songs on a 32GB storage device. Sounds feasible, doesn't it? Unfortunately, MP3 is still the prominent way of distributing music. It almost sounds like inserting cassette tapes into a blu-ray player! Why would someone use such a low quality format on these awesome new players?

Think back to the first flash MP3 players. About six or so years ago I had a Creative player with a whopping 32MB of storage, with the ability to insert an SD card, up to 512MB if I remember correctly. Now, 32MB is tiny by anyone's standard. Using the 3MB MP3 figure above, that's a measly 10 songs. It was important to have highly compressed music, lossless was not feasible. The trade-off was between quality and file size, for the ultimate convince factor. Why carry a bulky CD or Cassette player if you can have a compact solid state device?

With the advent of these small capacity MP3 devices came the distribution of music digitally. P2P was (and still is) a huge deal. Broadband had measly quotas, high costs, and limited bandwidth - assuming you were lucky enough to obtain it! Many people were still stuck on 56k dial-up connections. As with cassettes and CDs, piracy was not uncommon. With the Internet and P2P came sharing of copyrighted music, however this music had to be compressed considerably to facilitate those on low speed connections. The age of convenience was born. Any music could be obtained freely - on demand - and placed onto a portable compact device. However, this seemingly perfect concoction had a downside - music quality. 128kbps MP3 is by no means high quality. On good audio equipment, there's a distinct difference between it and lossless. For people with mediocre headphones, you probably can't tell the difference. To make the point easier to understand, here's an example of the compression artifacts introduced when lossy encoders are used:

Note: The author uses a .wav file, which is a lossless format.

I tried this with my FLAC copy of '1,000,000', a track by Nine Inch Nails. Here's the MP3 compression artifacts at 128kbps, and at 320kbps. Audacity and LAME were used to encode the MP3 files for this test. You can get the original track for free at NiN's website.

Despite the reduced quality, majority of people were happy at this stage. All the boxes had been checked, and the future of miniature digital players was underway. Audiophiles sourced lossless versions of their music to match their expensive high fidelity gear, so MP3 compression was not a concern. All was not perfect, though. You can choose a lossless music source, but you can't choose the master copy; and the music industry was starting to do some nasty things to their masters.

The Loudness War

There's a highly sought after quality in music called dynamic range. Dynamic range describes the ratio between the loudest undistorted signal, and the root mean square amplitude. In laymans terms, dynamic range represents the difference between the 'loudest' and 'softest' sound signals. If you've ever seen an audio waveform, you would have noticed that the higher the wave peaks are (i.e. the amplitude), the louder the signal is, relative to the waveform as a whole. A higher dynamic range means the difference between these loud and soft sound signals is greater, thus allowing a crisper reproduction of sound. When the dynamic range is limited, soft sounds come across louder, and loud sounds start to lose their punch. The process of limiting dynamic range is known as dynamic range compression (DRC). Think of it as forcing all the loud and soft sound signals into a smaller range of volume. The higher this compression is, the harder it is to differentiate between these loud and soft sound signals, because they approach the same volume. Exccessive use of this compression technique (with amplification) can cause clipping, which loses valuable audio data, and distorts the sound significantly.

When you're listening to music, chances are you'll want some insturments to have a strong kick, such as the beat of the drums, and other to remain soft and mellow. DRC kills this.

The record industry decided that the Compact Disc (CD) method of audio distribution would be a great opportunity to take advantage of dynamic compression. (As a side note, dispite the term 'compression', there's no change in file size when it is applied, unlike MP3 compression.) It might sound odd (no pun intended) that someone would want to apply DRC if there was no advantage to doing so. It just so happens that the definition of 'advantage' depends on what side of the story you're on. While the consumer may prefer a higher quality audio track, the industry thinks otherwise. Apparently people prefer their music made artificially louder. To demonstrate this, I've obtained a video which describes how sound engineers 'enhance' audio:

When a publisher has their music aired on the radio, they want it to stand out. Have you ever noticed how loud TV ads seem to be in comparison to the movie or show you were watching? That's DRC in action. They want to grab your attention - music is no different.

There's many examples of how this can negatively affect sound. Here's an example of DRC on Nirvana's hit 'Smells like Teen Spirit':

A great (and somewhat laughable) example of how bad this phenomenon has become can be seen by comparing Metallica's CD album 'Death Magnetic' to the Guitar Hero version. From what I can gather, Guitar Hero requires a decent amount of dynamic range in order for the music to function well with the game's audio analysis algorithms. This video compares the CD and Guitar Hero versions:

That's quite a bit of difference.

Why people don't care

Audio is often neglected by the general consumer. It's almost considered insane to purchase a $100+ set of high fidelity earphones, despite the massive difference in quality when compared to the likes of the common iPod offerings. Musical finesse rarely crosses peoples minds at all. You buy a CD or MP3 album, a new iPod, and off you go. No one questions the ability of the music player, let alone the earphones. Nor do consumers consider how the master of their favourite songs may have been sabotaged. When you're accustomed to listening to music from P2P clients, torrents, and YouTube, the artifacting introduced at low bitrates doesn't phase you. It's considered 'normal' for music to crackle and go wacky at high volumes, and having the loudest speaker system with the greatest amount of bass is considered 'audio bliss'.

Fortunately, for people who do care, there's a database which lists the dynamic range quality of many artists. It's called the Dynamic Range Database, and it's very handy if you want to avoid buying low quality media. If you want to test the dynamic range of your own tracks, the appropriate analysis software can be found here.

As it stands, CDs remain the easiest way to obtain lossless music. For people who want the very best, vinyl is still the way to go - according to extreme audiophiles. That being said, you need very high quality kit to hear the difference. This is assuming the master hasn't had a dose of DRC beforehand.

It's a sad state of affairs when it has got to the point where the masters are completely ruined to compete in a silly loudness competition. Perhaps there will come a time when having the prettiest MP3 player is considered a lesser priority than using decent headphones and rich high quality music. At least then people will come to appreciate just how stark the quality difference is between a 'good' and 'bad' track. We live in a world of ever advancing technology, let's utilise it! :)

For those interested, here's a few links to interesting articles relating to DRC use over the years:
What Happened To Dynamic Range? - Bob Speer
Wikipedia - Loudness War
Whatever Happened to Dynamic Range on Compact Discs? - George Graham

Note: I'd like to point out that I'm not a professional sound engineer. This article was written based on my own personal research. I've provided the sources for this article within.