As we learned in section 1.4, analog audio, in its electronic form, consists of voltage fluctuations going up and down to represent the sound wave. And in section 1.5 we learned how digital audio consists of binary code that the computer can translate through programming and algorithms to represent a sound wave.
The process of of transforming this analog audio to digital audio is called “ Analog to Digital Conversion and Digital to Analog conversion” or often by its acronym “AD/DA”
The AD/DA is a very important step in the process. This is where the bit depth and sample rate of the digital audio is determined, and the quality of the individual components have an impact on the sound quality by how accurately they convert the sound from one format to another.
The individual components of a converter are the following
- analog circuitry
- converter circuit
- word clock
- The analog circuitry on the AD side is fairly simple. The balanced signal needs to be unbalanced and buffered to the optimal impedance with a transformer, or a solid state circuit, and then it is fed into the converter circuit. On the DA side, the analog signal coming directly out of the converter circuit is low level, and therefore needs to be amplified to reach line level, and then balanced through a transformer or solid state circuit. The main analog circuitry components will consist of transistors, capacitors, op-amps and transformers. The quality of all these individual components will affect the sound quality.
- the converter circuit. This is the actual part that does the conversion. On the AD, it senses the voltage of the incoming signal with great precision, and takes incremental snapshots of this voltage and assigns a numeric value that corresponds to it. This numeric value is now in the digital realm, and in its root element is a sequence of 0’s and 1’s, called binary code.
On the DA side, the converter receives the binary code, and outputs the specific voltage that it represents. It does this repeatedly at a very fast rate, specified by the sample rate.
- Word Clock. In both the AD and the DA processes, the sample rate is what determines how often the converter circuit will take a sample of the audio, and convert it one way or the other. This sample rate, which is typically 44.1 kHz or 48 kHz, needs to be very precise, and is regulated by the word clock.
Just like if you have 2 wristwatches, and you synchronize them, after a year, they will likely be out of sync from each other by a few seconds due to a margin of error. Well even the most precise clocks in the world have a margin of error, and in audio conversion, this margin of error results in a loss of sound quality. Speaking of the most accurate clocks in the world, some high end audio converters are up there, by sensing the extremely steady decay rate of a nuclear source to control the clock rate. More commonly though, clocks will use a quartz piezo electric oscillator. Basically, the clock is a glorified metronome that dictates the precise timing of when to take a sample and convert it.
An important consideration when using multiple separate devices for audio conversion tasks, is that the word clock is synchronized on all the units, so the timing of when to take the samples is precisely the same across all units. Just like two wristwatches will be out of sync after like a year, two audio word clocks can be out of sync within less than a second, and the result will be audible pops and clicks. This is the reason many interfaces and audio converters will have a word clock in/out. Where there are multiple units that perform conversion, one unit will be chosen as the master, and the others as slaves. The word clock signal will be sent from the master to the slaves, and they will all remain in sync. There are several ways of sending this word clock signal, though a 75 ohm word clock cable is standard. The AES/EBU and ADAT protocols for digital audio are also capable of embedding the word clock signal within the digital audio signal, eliminating the need for a separate word clock cable.