Superior Performance Hi-Fi Products

Music On The Computer

Digital Audio Is Confusing!

Whether you're new to digital audio or have been using digital audio for a while the chances are you're finding it confusing. Just what do all those technical things mean and what use are they?

Overview Of Digital Audio

Digital audio is simply analogue audio that's been sampled to make it into data which then becomes more resilient to the corruption that analogue audio usually suffers from.

Whereas vinyl or analogue audio tape has to be protected a number of ways to prevent irreparable damage, a series of 0's and 1's recorded on media such as tape (DAT), a computer hard disk or a mem-stick, has a better chance of survival - it's more resilient. You can make any number of copies of data and copies of copies and they will still retain the fidelity of the original data.

Most people first experienced digital audio with compact disc (CD). Compact disc is not as resilient as it was first made out to be. It can retain data indefinitely but like vinyl it suffers damage during handling which can make it impossible for the CD transport to read it.

A general purpose computer can be used in place of a CD player. Instead of playing the CD it can be used to extract the data and place it on the hard disk so it can be played from there, or you can make back-up copies, or even copy the data onto a memory stick and copy it onto another computer hard disk.

Using a computer as your music source means you can also put music data onto it by downloading - by buying music online and transferring its data from a website to your computer. Because it's data there is no loss of fidelity of the data - your download is an exact copy of the original sample.

You can even copy your music data from the computer to another type of computer called a smart-phone, and use your smart-phone as a portable music player.

Playback Equipment

Although computers and smart-phones convert the digital audio data back to analogue audio so you can hear the music using speakers or headphones, the quality of music reproduction may not be to your liking.

This is because the computer device has so many functions that it cannot specialize in high quality audio reproduction. Even if the converter which converts audio data to an audible analogue signal (the DAC) is the best the manufacturer could afford to include, the computer device is not an ideal environment.

Sound Cards & External DACs

Soon after CD came on the scene the external Hi-Fi DAC followed hot on its heels. Why? Because the external DAC offered the user better fidelity. How? Because the external DAC 'lived' in its own environment (case) with its own power supply and featured the 'best' chip-set, often better than the CD player's own chip-set. All you needed was a CD player that had a digital output.

The digital output is known as a S/PDIF output which contains the digital data and the clock signal to which the digital data is synchronized.

They came in two flavours: coaxial cable (real wire) or optical cable ("TOS-link").

A basic computer doesn't have a S/PDIF output. To get that you need a particular sound card that has a S/PDIF output. Most basic computers have a sound card but more often than not they only have analogue outputs such as 'line' and 'headphones'. They will also have an input called 'mic'. These are basic sound cards. To drive an external DAC - the type made for CD players - the sound card must have a S/PDIF output.

There is a slight problem. Sound cards with S/PDIF outputs are getting quite rare today. Why? Because they're not that popular anymore, and big manufacturers only do things for monetary gains. They're not popular because a thing called USB (Universal Serial Bus) has become a more commonplace method of connecting computers with peripherals.

So now, to get a S/PDIF output from a computer you need a USB sound card. A USB sound card doesn't fit inside your computer like a traditional sound card, it's an external device.

USB Bit Depth and Sampling Frequency

When faced with the purchase of a USB sound card the customer is bombarded with a considerable amount of technical jargon urging you to buy product X and not to buy product Y.

A great deal of hostility exists not only between manufacturers but also between customers of one 'flavour' and another. They're fighting each other over numbers that have virtually no bearing on the actual quality you'll get!

Bit Depth

For USB audio there are two bit depths: 16 bit and 24 bit.

16 bit is what people call CD quality. It has a dynamic range of 96dB (decibels) in theory and that figure is often used in specifications for 16 bit audio. However, that omits the fact that the data has to be turned back into analogue for us to hear the music. The analogue section contributes some additional noise and so the practical noise is -90dB which makes the dynamic range 90dB.

24 bit isn't actually 24 bits of audio. 4 bits are used for non-audio data, so it is in fact 20 bit audio. In theory 20 bit audio is capable of 120dB dynamic range. In practice because of the analogue circuit and the noise contribution that makes, it is more like 105dB. However, manufacturers often quote 120dB although in practice that's never achieved.

Many believe that 24 bit results in better resolution than 16 bit. Advertising would suggest that. However, it is a complete fallacy! The only difference between 16 bit and 24 bit (20 bit in reality) is the dynamic range.

What that means to you is that the noise floor of 16 bit is -90dB and the noise floor of 24 bit is -105dB. Neither of which is audible when listening to music - even listening loud.

16 bit or 24 bit, you're going to hear the same sound quality provided all other things are equal.

Sampling Frequency

Sampling frequency otherwise known as the Nyquist/Shannon frequency after the names of the pioneers of digitization, is based on the theory that the highest frequency that can be sampled is half of the sampling frequency. It's a theory that's been proven to work.

The highest audio frequency a human can hear is 20kHz and that's a young human being. By the time you reach 40 you're lucky to be able to hear frequencies above 13kHz.

Therefore a sampling frequency of 40kHz is all that's required. There is no such sampling frequency as 40kHz but there is 44.1kHz (for CD) and 48kHz for professional studio digital tape recording.

There are people who believe they can either hear or sense frequencies above 20kHz. If this were possible they would be uncomfortably aware of numerous ultra-sonic devices and ultra-sonic emissions which exist in everyday life, but that is not the case.

What it could be is their equipment not resolving audible high frequencies properly and producing a distortion known as aliasing which 'folds down' ultra-sonics to interfere with audible frequencies. This can be avoided by good design.

Even so, some manufacturers have found offering higher sampling rates to be a good niche business which caters for this belief.

Therefore you will find sampling frequencies up to 96kHz on some USB devices.

Whatever the sampling frequency, from 44.1kHz upwards they all manage to reproduce 20kHz just the same. 44.1kHz reproduces 22kHz perfectly and 48kHz reproduces 24kHz perfectly.

Up-sampling And Over-sampling

The only problem in recording high frequency sound higher than one half the sampling frequency is it results in aliasing which folds down frequencies above one half the sampling frequency and superimposes them on the audible frequencies. This results in a sizzling or hissing high pitched sound similar to sibilance.

Today's A-D (ADC) chip-sets use up-sampling which still uses the 44.1 or 48kHz sampling frequency multiplied a number of times to move it away from the audio signal. This allows the audio signal to be attenuated above half the sampling frequency, at a rate that can be accommodated by analogue audio techniques, down to a level where aliasing distortion is so low as not to be a problem.

On replay D-A (DAC) chip-sets use over-sampling which moves the 44.1 or 48kHz sampling frequency upward in frequency, often by 8 times, so that the analogue output filter can have a gradual slope thus avoiding unnatural phase shift of the high frequency components of the music.

Last Word On USB 24/96

Even with such technology being available for USB there will be people who insist on 96kHz and above although the results are no different but their implementation can be much worse. For USB most 96kHz sound cards are based on experimental and often temperamental chip-sets in equipment with origins in the far east. You'd expect lower noise with 24 bit but often such devices are more noisy. What you won't often find is anybody admitting their 'big numbers' device wasn't worth the money they paid - it goes against the grain.