So what is digital sound?
After spending a weekend immersed in music at the local Truck Festival this year, music was still ringing in my ears, so I thought I’d write this post on digital sound. We find digital sound in everything from movie and music production, to podcasts, gaming sounds and the subtle interactive click of a button. The following is a condensed version of how sound travels from its original source, into the digital realm, is edited and played back again.
What is sound?
Sound is essentially a disturbance in the air that is created from an initial source, it could be the impact of hitting a drum, of strumming a guitar or a voice. Sound waves travel out from the source and eventually reach a receiver, this could be a microphone or an ear. Its only when it reaches our ear or a microphone and it is ‘perceived’ that it can be called a sound.
Each sound is different, and produces a different waveform pattern. The different waveform is what gives a sound its character, how we can tell a cymbal crash from a human voice.
What ‘shapes’ sound?
Two things affect how we hear a sound ‘Amplitude’ and ‘Frequency’.
Amplitude is perceived as ‘loudness’ or the strength of the sound and is measured in decibels or dB. 20dB is a whisper, 60dB is a conversation, 110 dB is the sound of a chainsaw and 130 dB is the level where sound starts to get painful.
Frequency the higher the frequency of a sound, the higher the pitch. Frequency is measured in cycles of sound waves per second or Hertz. 1 Hz = 1 full cycle of a sound wave per second, 1 kHz = 1,000 cycles or a soundwave per second, 6kHz = 6,000 cycles per second etc.
We can only hear a range of between 20 Hz and 20 kHz, below this range is sub sonic and above this range is ultra sonic. Dolphins and bats can hear well into the ultra sonic range picking up frequencies up to 200 kHz. Most of us don’t even hear up to 20Hz and as we get older we start to hear less than that.
In audio production this range is broken down further to lows (20Hz – 250 Hz), mids (250Hz – 6KHz) and highs (5kHz – 20 kHz).
Frequency response is the range of frequencies that can be perceived, whether that is by audio equipment or with the human ear. Audio equipment can come with a frequency response chart, essentially the flatter the frequency response chart the more accurately we can assume the device is able to record sound.
What is a note, and what is a sound?
A note is a pattern of sound waves that repeats itself over and over again, within the sound frequency that we can hear. Even the sound of a drill or a motorbike can be a note as long as it repeats itself. A sound is a single sound wave that doesn’t repeat like a door shutting, or a lock turning, it that sound were to repeat itself it would become a note.
How do we capture these changes in air pressure and turn it into digital signals, we need something that is sensitive to these air movements, for us its our ears, for digital sound its a microphone. The microphone takes one form or energy in the form of sound waves (acoustic) and turns it into electrical voltages (analogue) and this is then given a numerical or binary value (digital sound). This is known as analogue to digital conversion. Once the sound is in digital form, we can edit and manipulate it through audio software. Once we have manipulated it how we wish, we need to convert it back to an electrical voltage and back out through a speaker to push the air again.
Converting analog to digital
To take sound and use it in the computer we need a method to capture data this is done by ‘sampling’ sound. A sample is an individual piece of audio information captured at a point in time, capturing the frequency and amplitude of a sound wave. The more frequently we sample a sound, the more accurate our recording will be.
Sample rates - are expressed in Hertz, or frequency of samples per second. e.g. 44.1 kHz means in one send a sound is samples 44,100 times. The sample is then recorded as numerical data. 44.1 kHz is considered the minimum sample rate to achieve high quality audio, you can usually detect degraded quality below.
Bit Depth – a bit depth is the audio equivalent of shades of grey in a spectrum of white to black, it determines how many different shades of volume you can pick from, so 8 bit means 8 discreet volume intervals to choose from, as we increase the bit depth we can provide a greater ‘dynamic range’. Imagine the difference in sound intensity between the buzz of an insect and an explosion, and all the shades of sound in between, from conversation to music and birdsong, that kind of gives a picture of what we might be trying to capture, and with a greater bit depth we have a better chance of capturing the full dynamic range.
These two factors working together determine the quality of the sound, the higher the sample rate and bit depth the higher the file size. Cd’s are generally 16 bit (dynamic range of 96 dB) DVD’s often use 24 bit (144 dB) to cover a greater dynamic range.
The final stage of the digital process is audio delivery. We need to ensure we supply a balance between the right file size for the device the sound is being played on, and the quality of the sound. If sound is being supplied over low bandwidth internet, the consideration are completely different if you are supplying sound on a DVD. When we are working with sound we will be working with audio in an uncompressed format.
.WAV (Waveform Audio Format) – digital audio standard for Microsoft
.AIFF (Audio Interchange File Format) – digital audio format developed by Apple
SD2 (Sound Designer 2) – digital audio format developed by Digidesign
MP3 (MPEG-1, Audio Layer 3) – The Motion Picture Experts Foup format
WMA (Window Media Audio) – Microsofts compressed file format
AAC (Advanced Audio Coding) – Apple compressed file format
Just to give an idea of the difference audio compression makes, here are the sizes of a cd quality audio tract, and a compressed .mp3 file.
44.1 kHz at 16 bit data – 3 mins (CD quality) = 30 Mb
.mp3 file, at 128 kbps – 3 mins (good quality) compresses to = 2.8 Mb
Digital Editing Packages
There are a number of different audio packages to edit with, top end are tools like Protools and Cubase allowing for 8 track recording and professional sound production. Most video editing packages like Premiere and Final Cut Pro have the ability to edit sound along with video. Packages like Reason and Ableton Live are geared more towards digital music production, and there are some free simple audio editing software packages like Audacity that runs on Mac and Windows.
Also a great resource for learning about sound and where some of this information was sourced from is here at the Digital Audio Principles – Linda Online Tutorials.
What audio editing packages do you use or recommend? Which artist do you think has had the biggest influence on electronic music? Digital versus vinyl, now there’s a can of worms?