Related Articles

In the past couple years, the new standard in digital audio has become that of 24bit bit depth and 96kHz sampling rate (a CD, for reference, if 16bit / 44.1kHz). I am not one to partake in the analog vs. digital debate, but I will mention that this new standard represents a marked increase in quality when compared to the old 16bit/44.1kHz standard. A quick explanation is in order.

The bit depth is the number of bits that are used to represent the amplitude of an audio signal. The greater the bit depth, the greater the amount of steps between zero volume and maximum volume. With a bit depth of 1bit, there would be two steps in volume: on and off. As you could well imagine, that wouldn’t sound very good. This is very similar to how video cards deal with colour depth. At 16bits (the standard of several years ago), your video card is capable of representing over 64 thousand colours. The same can be said for digital audio – 16bits equals over 64 thousands steps in volume which produces a dynamic range of roughly 96dB. 24bit, on the other hand, equals over 16 million which produces a dynamic range of around 106dB. Quite an increase indeed! Why is this important? Because in “real-life”, our eardrums are analog devices so we hear don’t hear in “steps”. With high quality, digital audio, however, we can record digital audio that sounds good enough to trick our ears into thinking it is a real sound.

The sampling rate follows a similar train of thought except it represents range of frequencies of an audio signal. Simply put, our eardrums can distinguish frequencies as low as 20Hz and has high as 20,000Hz. The higher the sampling rate, the greater number of samples (or, audio “chunks”) taken per second, and the greater the resulting quality. The maximum frequency that can be represented digitally is equal to exactly one half of the sampling rate, ie: 44.1kHz maxes out at roughly 22kHz. (Why does this calculation work? Well it gets pretty complicated. If you want to know more, do a search on Google for +nyquist +"sampling rate" +audio). As you have probably already figured out, 96kHz can represent frequencies up to around 48kHz.

Why is this important if we can only hear up to 20kHz? In the world of audio, it is essential to maintain the quality of the audio signal until final output, or mixdown. Despite what we imagine, computers are not perfect as they produce rounding errors all the time. These errors occur all the time when dealing with complex digital data. When you are paying with audio everything you do – from applying FX to saving a file even – produces its own share of rounding errors. By working with very high quality audio data that is generally beyond what we humans can hear (ie: 24bit / 96kHz), we reduce the noticeable effects of rounding errors. (Hey science buffs, remember sig figs, or significant digits? In terms of audio, you want to keep as many significant digits as you can throughout the recording, mixing, and mastering process. More sig figs means less rounding errors means better quality audio.)

The topic of digital audio quality – specifically the need for 24bit / 96kHz and the age old digital versus analog debates – unearths deeply rooted opinions and is far beyond the scope of this review. At any rate (no pun intended), enough with this boring technical stuff. Let us move on.

send article   hardware newsletter   article comments (6)

Submit our article to: Digg this! del.icio.us