The Analog to Digital conversion process.
Hi ! This is Rafa Monteiro, from Montpellier, France. This is an assignment for lesson for week 2 of “Introduction To Music Production” at Coursera.org. Today, I would like to talk to you about the conversion of an analog signal into digital information.
Sound is a physical phenomenon. When sound happens (and it happens all the time), the mechanic waves travel across the air, ground and water in a continuous way, with the intensity of the sound becoming smaller the further it gets from it’s source.
When the sound hits a microphones or an electric pickup, from a bass or guitar, what it does is create an electrical signal that is analog to the sound in the ambience. Since it’s analog, it’s also a continuous signal, with an amplitude that relates to the intensity of the sound, a frequency relating to the number of sound vibrations, and the decreasing intensity in time, just as the sound does.
A computer, unfortunately, can’t understand sound or electric signals in it’s nature. This happens because computers don’t understand continuous information. They only work with binary data, composed of huge sequences of information written only in zeros and ones – that’s why it’s called binary information.
All data processed through a computer is written in big chains of zeros and ones. All programs, all internet content, this text, all the videos and musics played on a computer. It’s no different from the recording process.
In order to work with the audio signal in a DAW, the signal must be transformed into binary data.
This conversion is made in the audio interface, in a device called “AD/DA” converter, which means “analogic to digital/digital to analogic” conversion. This electronic component will read the audio signal and write binary words that are analog to the signal. Those big words, full of data, can be understood by the DAW.
This conversion is made through a quantization of the signal: periodically, the device will capture discrete amounts of information from the signal to build the digital data.
This data is build on bits. A bit is a piece of memory that can represent either zero or one. A word of one bit can only represent two values: zero and one. A word of two bits can represent more values: 00, 01, 10, 11. three bits can represent even more data: 000, 001, 010, and so forth.
Since sound has a lot of information and nuances that need to be registered in a high quality recording, the digital words used to store this data are very big. Home CD systems use 16 bits of information only to store audio information, and for studio standards, it’s usually in 24 bit word information. This is called the resolution of the recording, and it’s related to the amplitude of the sounds recorded.
Frequency is registered in another way, by the number of samples made during the conversion of the audio signal into digital data. For each second, the AD converter will dig samples of the signal.
As you can see in the graph above, a good quality in audio demands bigger samples. In order to record all the possible different audio frequencies (they are between 20 Hz and 20k Hz, remember?), it’s necessary to use a very big sampling rate. For home CDs, it’s 44,1 kHz sampling rate and for professional studio, it’s usually 48 kHZ or even 96 kHz.
That also means that the equipment you are using will be burdened when processing these tons of data. Even small recordings, made of two or three instruments, can generate insanely huge amounts of data. Usually, home computers won’t be able to handle all this processing smoothly, and that’s why the purchase of a dedicated DAW should be considered and studied.
Last, but not least, the quality of the AD/DA converters differs from one audio interface to another. Better quality AD/DAs usually means more expensive equipment, but also a better quality of digital audio. As I explained in my previous article, the quality of the sound is usually given by the weakest link in the chain of equipment, and that’s another topic to consider when purchasing an Audio Interface.