Sample rate and bit depth each respectively affect a huge aspect of the audio you’re recording and working with. Let’s give a brief overview of the two and most importantly identify what sample rate and bit depth should I use.
What Are Sample Rate and Bit Depth
Sample rate and bit depth affect the frequency and dynamic range of your audio, respectively. Let’s give a brief overview of each to better understand what we’re changing when we change the setting for each.
What is Sample Rate
Sample rate refers to the frequency range you can capture and represent with your audio. A sample rate of 44,100Hz is just that; it means that the highest frequency able to be reproduced in that audio is 44.1k.
You may have heard me state in the past or already know that that the human ear is only capable of hearing up to 20kHz at most, leading you to wonder why we would need a sample rate much higher than that?
There’s a curveball explained in the Nyquist–Shannon sampling theorem which states that in the digital realm, the sample rate must be double the bandwidth of the signal to avoid aliasing.
As I explained in my overview of what is aliasing, this is a form of distortion which results from a fold over error created in trying to play frequencies higher than our sample rate is capable of reproducing.
So in actuality, a sample rate of 44,100Hz is only able to reproduce frequencies up to half that, or 22,050Hz in digital audio.
Another popular sample rate of 48kHz reproduces frequencies up to 24,000kHz in digital audio, a sample rate of 96kHz reproduces up to 48kHz, and so on.
Now that we’ve given an overview of the frequency side of the coin via sample rate, let’s move on to dynamic range with bit depth.
What is Bit Depth
As I explained in my overview of bit depth, this relates to the amount of dynamic range capable of the audio. This is effectively the difference between the noise floor and the point of clipping, in decibels.
Every “bit” equates to 6 decibels. Audio with a bit depth of 16 has 96dB of dynamic range (16*6).
What is the industry standard in bit depth for audio production, 24 bit, has 144dB of dynamic range (24*6).
This means that there’s 144 decibels between the noise floor and the point of clipping to work with.
The more of a difference between the two, the cleaner your audio will sound because you can put more distance between your audio and said noise floor.
Bit depth doesn’t just relate to the decibels, but also values which the amplitude of a sound wave can be assigned.
The max “volume” of a sound wave varies from sample to sample, and each sample is assigned a value to refer to its amplitude.
Higher bit depths allow for more precise values to assign amplitudes to because they offer a far greater range.
Unlike the dynamic range, these values increase exponentially with the bit depth.
Specifically, a 16 bit depth means 2 to the 16th power in terms of these values. In other words, 16 bit depth has 65536 (2 to the 16th power) values to assign to the amplitude of a sound wave from sample to sample.
Because it’s an exponential jump, 24 bit depth has 16777216 (2 to the 24th power) values, making it far more suited for assigning the appropriate value to each amplitude.
When the bit depth doesn’t allow for the value that the audio’s amplitude calls for in that particular sample, it gets rounded to the nearest available value in the audio equivalent of trying to force a square peg into a round hole.
This results in a quantization error which manifests as harsh and unpleasant distortion not unlike the equivalent with the sample rate aliasing error.
This is a fact of life when we convert the bit depth of a track from 24 to 16 bit for CDs (16 bit being the max bit depth the medium of CDs can reproduce).
To “fix” the quantization error distortion we get when we fit a seemingly limitless number of samples’ amplitudes into round holes, there’s something called audio dithering which replaces the distortion with low level white noise at the noise floor.
But I digress, let’s get back to the topic at hand, what sample rate and bit depth should I use.
What Sample Rate and Bit Depth Should I Use
Generally speaking and to answer what sample rate and bit depth should I use, 44.1kHz and 24 bit depth are perfectly adequate for capturing and reproducing your audio at full fidelity.
In fact, I spent the majority of the time talking about 24 bit depth and 44.1kHz sample rate earlier because they offer more than enough range for dynamics and frequencies, respectively.
Why the controversy then, or why is this a discussion to begin with?
Some audiophiles swear by higher sample rates such as 88.2kHz or 96kHz because when we add distortion like saturation to our audio in the mixing stage, it’s creating harmonics and at higher frequencies which didn’t exist before. These frequencies can exceed 22,050Hz (in the case of our 44.1kHz sample rate) and, although we can’t hear those frequencies, we CAN hear the artifacts of the aliasing this would create.
In the past, foldover errors which resulted in distortion were more prevalent.
Nowadays, something called oversampling is commonplace in virtually all modern plugins which create harmonics, meaning they increase the sample rate in the context of what they’re doing to the audio to raise the frequency threshold much higher.
They then downsample on the out and use an anti-aliasing filter as they pass the audio on to the next stage in the plugin chain so aliasing isn’t an issue.
32 bit depth is also a thing, but the huge array of amplitude values and dynamic range 24 bit depth offers is more than enough to capture even the loudest of live gigs, so it’s certainly good enough in the studio.
Just make sure to use my gain staging cheat sheet to make the most of the dynamic range you have and get the best sound from your audio.
In other words and if you’re smart about it, 99% of the time working with a 44.1kHz sample rate and 24 bit depth provides more than enough “headroom” both in frequency and decibels, and anything higher is just creating a needlessly larger file size.