Digital Audio
Ever since the rise of DAWs (Digital Audio Workstations) in recording we have had to learn about digital audio. How is digital audio different than analog? How is digital audio measured? How is analog signal generated in to digital audio? Why is it important to pick the correct measurements when recording? These are all questions that I had to answer for my myself and tweak my system to get the best quality after conversion.
How is digital audio different than analog?
Lets take a look at what makes up an analog audio signal. When you plug your guitar in your amp and play a note, if you are recording this with a microphone, then the mic will read a voltage.That voltage drives electric current to a speaker that then creates a waveform, which is made up of two parts. One is amplitude, which is how loud the note is, and two is frequency which is how fast the waveform’s cycles move per second, or as we perceive it as how high or low the note is on the scale. If we are using software to record this note then the computer has to take these two measurements and convert them into binary which is how a computer reads data. With analog signal the waveform is continuous. With digital audio the waveform is broken down in to samples or snap shots. These snap shots or samples have to be read in a way so they are continuous, so that when it is played back it sounds like an analog waveform.
How is digital audio measured?
Before we talk about how analog signal is generated into digital, we must know what digital audio is made up of. So if analog audio is made up by frequency, which is cycles per second, and amplitude which with audio is measured by decibels, and digital audio needs to recreate this waveform. Then it needs to have similar components that it is measured by. With frequency digital audio makes snapshots or samples for in each second of the wave form.The amount of samples a seconds are called sample rate. sample rates are labeled in hertz or kilohertz which is how frequency is also measured. What does this mean to the sample rate in my session? It means that if your sample rate is 44.1 kHz then there are 44,100 samples taken a second. The sample rates that are more popular are 44.1 kHz, and 48 kHz. Now digital audio uses quantization to determine the decibel level or voltage of a sample point over time, turning it into binary digits or what engineers call Bits. Bits are places on the waveform that the samples maybe round up or down to on the chart to specify a direct location. Bits are used then in DAWs to determining how many point of information can be saved in each sample over time, called bit depth. The bit depths that are more popular are 16 and 24 bits. 16 bit sample can hold up to 65,536 points of information, and a 24 bit sample can hold up to 16,777,216 point of information. There is a huge difference between the two.
How is analog signal generated into digital audio?
Analog to digital converters take care of most of this process, but how are analog signals represented in the non continuous binary representation. the wave form is sampled first to find the points on the waveform that need to be quantized. Then the A/D converter quantizes the samples to the closes binary digit. The more bit depth you have the better the quality of accuracy to the analog waveform. It wont matter how high your sample rate and bit depth is, it wont match the analog waveform perfectly, this is called signal to error ratio. Digital audio is measured on a graph in steps and these step make the wave form squared off, when the waveform is squared off it add very low distortion to the audio playback. we can actually fix this by adding a little bit of noise to the audio conversion, this noise is called dither. Dither helps reduce that distortion and improves the conversion.
Why is it important to pick the correct measurements when recording?
When recording it is all based on the interface you are using and the A/D converters in that interface. Most interfaces can go up to 96 kHz sample rate and 24 bit. The most important decision is made by asking yourself what format will my music be played on. If it is red book audio standard (CD Standard) then 44.1 kHz 16bit is what the audio needs to be converted down to. If it is DVD quality then it will be at 48 kHz 16 bit. The conversion between 24 bit to 16 is an even conversion so there is no loss in information it is just evenly condensed. This can't be said about all sample rates though. Lets say you record at 96 kHz and are going down to 44.1 kHz for CD standard then the conversion is not even, this could result in artifacts showing up in the condensed audio file. In this case it may be the best decision to record at 88.2 kHz for an even down conversion. These are things that I think about when recording and that mastering engineers like to see when converting down the audio to the format that it will be distributed at.
Knowing the basis of digital audio will let you make quality decisions when recording either at home or in a studio. I encourage you to research more about these processes to have better recordings.