Question: What is bit depth and sample rate?
Answer: In our discussion about high-resolution audio, we briefly touched on bit depth and sample rate. Here we want to talk about this in more detail. When talking about bit depth and sample rate we are talking about digital recordings of audio. Analog audio is converted to digital via an analog-to-digital converter by taking samples or snapshots of the audio at a very frequent and fixed rate (sample rate). The samples are also of a fixed size (bit depth). These two factors – sample rate and bit depth – weigh heavily in determining the quality of a digital recording.
Before talking about bit depth and sample rate, just a brief discussion about how computers handle data is in order. Almost everyone involved in audio today has spent most or all of their lives in the computer age. Unless you have been hiding under a rock somewhere, you probably have a fairly decent grasp on how computers work, so we’ll be brief. Computers look at everything as either being “on” or “off” or, as more commonly represented, as a “0” or a “1.” This is known as a binary digit (bit). So, every instruction given to a computer is provided in a string of bits. A “byte” is a grouping of 8 bits. A digital word is composed of one or more bytes. So when we refer to 16 bits, we are saying there are 16 digits (with each being either 0 or 1) that make up a word. Wow! Glad we got that out of the way.
So now, let’s talk about bit depth. The number of bits in a digital word determine just how accurate the information in the sample is. For a related example, suppose you ask someone to tell you about their new car, but can only use 20 words to do so. Then ask the same question, but allow them describe it using 50 words. Which of the two descriptions will provide the most detail about the car? The same applies to digital audio. Within this data is information about the audio at the time the data was sampled. The greater the bit depth, the more precise the information that can be collected (less rounding/quantization errors as bit depth increases). Just to give you an idea about the effect of bit depth on digital audio, in a 16-bit sample there are 65,536 different values. For every bit of resolution added the unique combination of values doubles. In a 24-bit system, that number jumps to 16, 777,216 different values. Remember we are talking about just a single slice or snapshot in time. Please note that CD-quality audio is 16-bit.
Let’s move on to sample rate. As already stated, bit depth refers to how much information is collected (sampled) about the audio at a given place in time, while sample rate refers to how often that information is collected (sampled). Another overly simplistic example is in order. Suppose your friend is at a basketball game and calls you every two minutes to tell you exactly what is happening at that very moment (and only at that moment). Now suppose he calls you every minute and tells you exactly what is happening at that very moment. From which of these two examples would you learn the most about the game? Again, this idea can be applied to a digital audio file. A faster sampling rate can provide a more accurate representation of the audio than a slower rate. There are some practical things to consider when selecting a sample rate. The Nyquist-Shannon sampling theorem states that the sampling rate must be at least two times greater than the highest frequency that is being sampled. Please note that the CD-quality audio is sampled at 44,100 Hz.
So, why not just use an extremely high bit depth and very fast sample rate at all times? Wouldn’t this ensure the best audio quality possible? Sounds good, but there are tradeoffs. First, higher resolution audio takes up a lot more storage space. It also takes a lot more system resources to record or play higher resolutions files and this may be perfectly fine in certain applications. If your goal is to ultimately place the recording on an audio CD, then the higher resolution file will need to be converted to 16 bit, 44,100 Hz before doing so.
Audio-Technica offers several USB microphones, such as the AT2020USBi and ATR2100-USB, that provide high-quality analog-to-digital conversion with a 16 Bit, 44.1/48 kHz sampling rate or better. If you have additional questions or would like a recommendation for an Audio-Technica product, feel free to contact the Audio-Technica Audio Solutions Department. We are always happy to help.