Soundware Beginners Guide to...
...Basic Digital Audio
To produce high-quality recordings on your computer, it helps to know a little about the theory of sound and how it relates to digital audio.
Comments/suggestions for our beginners’ section? Contact us!
- Sound Basics
- Sound into electricity (analogue audio)
- Analogue to digital conversion
- Resolution, sample rates and sound quality
- Digital audio formats
- Advantages of digital audio
Sound BasicsSound is transmitted through the air as waves of high and low air pressure created by a vibrating object such as a guitar string. The number of vibrations per second is known as the “frequency” of a sound – low frequency sounds sound low pitched to us, whereas higher frequencies are heard as higher pitches. Frequency is measured in kHz, and the accepted range of human hearing is of frequencies between 20kHz and 20,000 kHz – between twenty thousand and twenty million vibrations per second (although this decreases, particularly at the high-frequency end, with age or owing to damage caused by excessive loud noise).
Sound into electricity (Analogue Audio)To record audio, the sound first needs to be converted into an electrical signal. The process by which this is achieved takes advantage of the “motor” effect – the fact that when an electric current is run through a magnetic field, a force is created. So, if a magnet is suspended within a wire coil and a current is run through the coil, the magnet will move (or, if the magnet is fixed, the force will be transferred and the coil will move instead). Conversely, therefore, if a magnet is moved within a wire coil the opposite will occur – an electrical current will be generated in the wire.
So, to make a simple microphone, a wire coil is attached to a diaphragm (a flat piece of material that will vibrate like a drum skin) and a magnet is fixed within the coil. When sound waves hit the diaphragm it vibrates, and the spring moves at the same frequency as the sound waves. Since it is moving within the magnetic field created by the magnet, an alternating electrical current is created that is proportional to the frequency of the sound vibration that hit the diaphragm.
A simple moving-coil microphone.
To make a speaker, the process is reversed – the electrical current from the microphone is passed through another coil, again attached to a diaphragm and with a magnet fixed within the coil. The current in the coil within the magnetic field causes the coil to move, which causes the diaphragm to move – again at the same frequency as the original sound vibrations that caused the microphone diaphragm to move. As the diaphragm moves, it will cause the air around it to vibrate as well, and will transmit sound waves identical (in a perfect world – in reality there is always a certain amount of electrical interference, resistance etc. which affects the resulting sound) to the original sound waves.
A simple loudspeaker.
This is the principle of all analogue audio recording equipment and is the basic design upon which all microphones and speakers are modelled.
Analogue to Digital ConversionAnalogue audio works according to variations in an alternating electrical current. The differing voltages and the alternation of the current corresponds exactly to the original sound wave (the word “analogue” comes from the same root as “analogy” and “analogous” – referring to things that relate exactly to something else). However, digital technology works on completely different principles – data is converted into binary code made up of ones and zeroes. The analogue medium can express infinite variation in signal levels, whereas digital technology works on an “on/off” principle by its very nature (think of a digital watch with changing numbers compared to an analogue clock face with gradually moving hands.) So, to convert analogue audio into a digital format it must be converted into binary code that can be read by a computer.
To do this, the electrical current that represents a sound is passed through an analogue-to-digital converter. This takes readings of the voltage of the current many thousands of times per second, which it then rounds to a whole number and converts into a binary value. Conversely, a digital-to-analogue converter reads the code and converts each of the binary values to an electrical pulse that makes up a current very similar to the original analogue signal. These two converters are usually combined into one AD-DA (analogue-digital, digital-analogue) converter.
Resolution, sample rates and sound qualitySample Rates
To convert an analogue signal into a digital format, it is passed through an analogue-to-digital converter. This takes readings or “samples” from the electrical signal many thousands of times a second and converts the readings into binary code. So, the higher the number of samples taken (the sample rate), the more accurate a representation of the original analogue source is delivered and the higher the sound quality. The standard sample rate for CD-quality audio is 44,100 samples per second, which is written as 44.1kHz. The majority of professional digital audio devices operate at 96kHz or higher.
Low Sample Rate
At low sample rates, the gaps between each reading of the current are large, and the wave is not converted very accurately, resulting in lower sound quality.
High Sample Rate
At high sample rates, the gaps between each reading of the current are small, so a more accurate representation of the wave is produced, resulting in higher sound quality.
ResolutionThe resolution of a digital signal is the range of numbers that can be assigned to each sample. When the sound is passed through an A-D converter, the values that are “sampled” are rounded to a binary value. So, the greater the number of decimal points the number can be rounded to, the more accurate a representation of the actual reading it gives and the higher the sound quality will be. The number of binary values assigned to each sample is measured in bits (CD quality is 16-bit, and most professional audio devices operate at 24-bits or higher). Low bit rates can cause “quantization distortion”, where the sampled values are rounded up so much that they no longer give an accurate representation of the sound, and the digital audio signal takes on a crackly, metallic quality. Higher bit rates are also capable of transmitting a higher dynamic range, resulting in deeper bass and clearer high frequencies.
At a low resolution, the readings are heavily rounded, giving an inaccurate representation of the wave and a lower sound quality.
At a higher resolution, the readings are rounded less and give a more accurate representation of the wave, resulting in a higher sound quality.
Digital Audio FormatsDigital audio can be stored in many different formats. In professional audio work, uncompressed audio file formats are used as these generally give the best sound quality. Some examples of uncompressed formats are:
AU: A file format (abbreviation for "audio") that originated on the Sun and NeXT computer systems. Not widely used today.
Audio Interchange Format (AIF, AIFF): File format for Macintosh system sounds, similar to Windows' WAV format.
Compact Disc Digital Audio (CDA): This is format used for encoding music on all commercial compact discs.
SND: Another file format (abbreviation for "sound") similar to the AU format and used primarily for Macintosh system sounds.
Waveform Sound Files (WAV): This format (pronounced "wave") produces an exact copy of the original recording, with zero compression.
A problem with uncompressed digital audio is that the file sizes can be very large, however. To combat this, a variety of compressed audio formats exist that are mostly used to transfer audio over the Internet. These use “codecs” (short for “compression/decompression”). A codec converts an uncompressed digital audio file into a smaller format, and will also decode the compressed file enabling it to be played back. Some codecs work by physically removing data from an uncompressed audio file to make it smaller. In this case the file cannot be converted back into an uncompressed format from the compressed file. This is known as a “lossy” compression method. Others allow the original uncompressed file to be retrieved from the compressed file – these are known as “lossless” compression methods. Lossless compressed files are generally larger than lossy ones, though.
Advantages of Digital AudioThe debate over whether digital or analogue audio technology sounds better is long running and will doubtless continue for years to come. In truth, the difference in sound quality between top-of-the-range digital and top-of-the-range analogue equipment is almost indistinguishable. However, for the home user, digital technology has a number of benefits over analogue:
Value for money
Digital technology has made home recording accessible to a much wider audience. Computer multitracking software has in many cases removed the need for hardware mixing consoles and effects, and as such allows a basic recording setup to be much cheaper and more compact. In particular, software emulations of analogue hardware (such as vintage synths) are considerably cheaper than the real thing!
Digital recording is immune to much of the electromagnetic interference that causes noise on analogue recordings. This makes it much more suitable for home recording, as often home studios operate in less-than-ideal surroundings without any of the expensive electrical isolation and shielding of professional analogue studios.
Digital audio can be copied infinitely without any loss of quality, unlike analogue audio, which loses quality with each reproduction.
Digital media such as CDs are much more durable than their analogue counterparts.