When people talk about MP3 files and the like, you often hear the term Bit Rate also mentioned, so what’s it all about?
Until the mid 1980s, commercially purchased music was always in analogue format, be it typically vinyl or cassette. In late 1982, the arrival of the now ubiquitous compact disc (CD), co-developed by Sony and Philips, changed all that and revolutionised how we listen to music.
In order to create digital recordings such as CDs, the conventional analogue audio first has to be digitised. The process of digitisation takes the original audio source and chops it up into lots of thin slices. Each slice represents the volume of the audio source for a particular point in time. A digital number then stores the volume level of this slice on the CD. Each slice, or sample as it’s actually called, represents the level of the audio for a mere 23µs (23 millionths of a second).
To reconstitute the audio track, all of these samples are stuck together and the volume level of each sample is sent out to your speakers every 23µs (in the case of CDs). It’s a bit like slicing a loaf of bread and then squeezing all of the slices of bread together again to make it look like the original loaf. Not perfect, but if you squint, you’d never know that difference. CDs are very similar, because the slices are so thin, you don’t know the difference.
The rate at which we take the samples and replay them is called the sample rate. For CDs, 44,100 samples are taken every second! Now, because we have to store the value of every sample, it follows that higher sample rates would require a lot more storage, which can become a problem. Conversely, a higher sample rate directly equates to a more accurate representation of the original signal, and therefore better audio quality. So, ultimately compromises have to be made and a 44,100 sample rate was deemed to be a sensible compromise when CDs were developed.
Now we have an overview of sample rates, we are in a good position to discuss bit rates, as the two are inextricably linked. Each sample we discussed above is represented by a digital number and every digital number is made of digital bits; there’s the magic link! A bit is a single digit, either a one (1) or zero (0). In the case of our CD, each digital number comprises 16 bits, which allows numbers up to 65,563 to be stored; over 65,000 different volume levels!
As we have two ears, we need to store two sets of samples, one for each ear; stereo.
So, to calculate the bit rate for CD audio, we multiply our sample rate (44,100) by 2 (for stereo) and by 16 (the number of bits in our digital numbers):
44,100 x 2 x 16 = 1,411,200 bits per second (bps) = 1,411 kbps
We have now calculated the bit rate for the standard CD, which turns out to be 1,411 kbps.
If we used a 1 GB MP3 player to store music at the above bit rate for example, we would only be able to store 100 minutes of music; not very much! This is why MP3 was invented, as a way of reducing the amount of space that music needs when stored on a computer or music player by squashing, or compressing, the original CD audio data to save space. There are numerous compression systems available, with MP3 being the most popular, but they all aim to provide the same result; significantly reduce the audio file size.
How the compression works is somewhat beyond the scope of this article, as it uses a lot of complicated mathematics to work out which bits of the music we can hear, and which bits we can’t. Essentially it saves space by not bothering to store the bits of the music that we can’t hear. For example, if you have a Heavy metal band playing with the volume turned up to 11 and someone at the back of the stage taps on a triangle, you are unlikely to hear it. So MP3 decides that the tour bus is too full and fires the triangle player, saving space and reducing the queue to the bathroom at the same time!
During this compression process, typically 90% of the data is thrown away, leaving file sizes about one tenth of the original size, which in turn means that your MP3 player (or iPod for the image conscious types) is able to store ten times the number of songs compared to if it had to store the original uncompressed CD data.
Bit Rate Versus Quality
Now that we have reduced our file size to approximately one tenth of its original size, we only have one tenth the number of samples and so consequently only have to send them out to our speaker system at one tenth the original rate, i.e. our bit rate is ten times slower. A typical bit rate for good audio quality would be around 128 kbps (kilo bits per second, or thousand bits per second), approximately one tenth of the bit rate for the standard CD.
Unfortunately compression system have limitations, the more data they throw away, the smaller the file size, the lower the bit rate, but the worse the audio quality, or to use the band analogy above, more and more members of the band get thrown off the bus, so there are less musicians to make the music.
So, we should now have an understanding of what bit rates are and their direct correlation with audio quality. As a useful guide line for the quality of different bit rates:
- 320 kbps – Almost indistinguishable from the original CD
- 128 kbps – Reasonable audio quality, typical for MP3s
- 64 kbps – Not recommended for music, but good for speech
- 32kbps – Poor, often used to reduce web download times
Hear for Yourself
If you’d like to hear for yourself the direct effect of bit rates on audio quality, check out the following demonstration MP3 tracks, courtesy of Silicon Bay: