Digital Music - what, why and how
The University of Erlangen-Nuremberg might seem like a curious place to kickstart a revolution in how we listen to music but it was here throughout the late 1980’s that Karlheinz Brandenburg worked to develop the first widely available music compression system (the story goes that Brandenburg refined the process by repeatedly listening to Tom’s Diner by Suzanne Vega so he almost certainly deserves our sympathy as much as he warrants our respect). The result was the MPEG layer 3 compression system, or MP3 as we more commonly know it.
Today, a huge amount of music that we listen to is compressed. The reasoning behind compression is logical enough. Even when we can fit more data into less space than ever before there are desirable reasons for being able to make maximum use of that memory. This article is a short guide on compression types and formats and the best way to compress your music while keeping the quality levels up.
All compressed music discussed here is digital. At a basic level information is carried as bits and those bits are composed of ones and zeroes. If these bits can be seen as the “alphabet” of digital music, the various different formats or codecs can be seen as languages to encode and compress these bits. The measurement for how much compression is applied to music is how many thousands of bits a file uses in one second. The Kilobits per second (bit rate) calculation or kbps is the default measurement for the quality of a compressed audio file.
If we take CD as the “baseline” of music with compressed below this point and high resolution above, a song on a CD is sampled at 1411kbps. As you can see, even when we are talking about high quality compressed audio files, more than two thirds of the total information on a disc is removed in the compression process. Even so, 320kbps files can still sound uncannily close to their CD source if they have been compressed carefully. So what the available codecs for compressing?
MP3 was the first of all of these compression formats and remains the most popular choice to this day. There are several reasons for this. First, we are most familiar with it as a format because it has been around for a long time. We often use “MP3” as a short hand for compressed music even when it has been encoded in another format. MP3 is also incredibly widely supported. Not only is it pretty much universally accepted by players that read from hard drives (or have their own fitted), it is also frequently supported if you burn MP3 files to CD.
The format has evolved over the years as well. When MP3 was launched, the limitations of the encoding meant that you selected a bit rate to encode music to and the entire file was ripped at this level. More recent developments allow you to use a method called Variable Bit Rate recording. This allows you to select a bit rate that compression should be based on (for example 192kbps) and the encoding software will then study the file it is encoding. Complex parts of the recording will be ripped at this bitrate (and some parts at more than this) but where much less information is present, the encoder will reduce the bit rate to save on space. This means you have less compression when you don’t need it and more when you do.
MP3 is of course not the only game in town any more. The arrival of the iPod and supporting iTunes introduced another compression format to the public. Although generally associated with Apple, AAC (Advanced Audio Coding) was originally designed to replace MP3 but hasn’t quite succeeded in doing so yet. AAC uses a more sophisticated encoding system and a general rule of thumb is that any bitrate of AAC will be audibly the same as the next bitrate up of MP3. This does not always hold true but it is fair to say that AAC is more efficient than MP3. Although AAC is less widely supported than MP3 (especially with regards to burning it to disc), it is still common enough to be a very logical choice and Cambridge Audio equipment is as happy with AAC files at it is with MP3. Historically, the only piece of software that was commonly available to rip CD’s was iTunes but if you aren’t a big fan of Apple there are some third party pieces of software that will also encode to this format.
The long standing rival to AAC is Windows Media Audio (WMA). This has been around almost as long as people have been compressing audio but because it is essentially a Windows tie in, is less commonly used than the other two codecs. Like AAC, the software behind WMA is more sophisticated than MP3 and the same general rule of thumb about compression holds true. If you are a big user of iTunes, WMA is essentially a no-no (it will ask to convert it every time it detects a file which can become a little wearing) but if you are a happy Windows user, it is an effective codec.
Cambridge Audio naturally believes that the best way of listening to your music is uncompressed and we cover the ways of getting the best from your lossless media in an upcoming article. If you do need to use compression though, there are some simple guidelines to follow to achieve the best of both worlds. The first is that although all the codecs will support bitrates as low as 32kbps, you really don’t want to go that low! Remember that a CD file is 1411kbps which means that these very low bitrates are only 2-3% of the original. The results of this will definitely be audible (a really easy test is to listen to live applause- at these very low rates it becomes a noise rather than the defined sound of people clapping). You can fit a very large number of albums into not a lot of space doing this but there is a good chance that you won’t want to listen to them.
Choosing bitrates of 192kbps and above should produce results that save a large amount of space in storage but also keep most of the musical information intact and the higher you can go the better. With storage now much more cost effective than it was, ripping at 320kbps should reduce the size of an album by about three quarters and even large audio collections will fit on a single external hard drive. If you are using MP3, it is also best to use variable bit rate encoding.
A more recent form of compression is Lossless compression. This might sound like a contradiction in terms but the idea is to take a CD sized file and apply a process derived from more conventional compression to reduce the size of the file without affecting the quality of the sound. It would be easy for a person or company to make bold claims about how far this process can be taken so there is a simple benchmark for lossless compression. In order to be defined at “lossless” it must be possible to exactly recreate the original uncompressed file from the compressed one. If you can’t, it isn’t lossless. We will cover lossless audio in due course but the important thing to remember is that lossless files might be shrunk from CD but they are still much larger than the compressed type codecs and will require more space.
To summarise; when compressing your audio, think about the following:
- Where are you going to use your compressed audio? This will help decide what format is best for you.
- Don’t over compress your music- the results will be unlistenable.
- As storage is relatively affordable, choose your storage to fit your music and not the other way around.
- If you are considering lossless compression, your storage requirements will be much higher.
- Taking the time to tag your music when you rip it is the easiest time to do it.
As with so many parts of audio, the best judge of what works best is your own ears. If you take our thoughts on compression as a benchmark, run some tests of your own and see what format and bitrate is most suitable for you.