Frequency State / Music Production Reference Library · Volume 1
Sample rate determines how frequently a digital audio signal is measured; bit depth determines the number of quantisation levels available to store each PCM sample. They affect different technical limits, and neither should be treated as a simple quality score.
The two main settings discussed here are sample rate and bit depth. In uncompressed PCM audio, they describe when measurements are taken and how each measurement is represented numerically.
Together, these define the quality and resolution of digital audio.
What is Sample Rate?
Sample rate describes how often audio is measured per second.
Sample rate is expressed in samples per second, commonly written in hertz or kilohertz. A 48 kHz recording contains 48,000 samples per second in each channel.
Common sample rates include:
Common rates include 44.1 kHz, associated with audio CD, and 48 kHz, widely used for video and media production. Higher rates such as 96 kHz are also available, but whether they improve a particular workflow depends on conversion, processing and delivery requirements.
A higher sample rate captures more detail in the sound.
How Sample Rate Works
Digital audio captures snapshots of a sound wave at regular intervals.
More samples per second means:
A higher sample rate extends the theoretical frequency range a PCM system can represent, subject to adequate anti-aliasing filtering. It does not simply make every audible waveform ‘more accurate’ or guarantee audible improvements.
The Nyquist Principle
To accurately reproduce a sound, the sample rate must be at least twice the highest frequency present.
For example:
For a 44.1 kHz system, the Nyquist frequency is 22.05 kHz; 48 kHz permits an ideal Nyquist frequency of 24 kHz. These rates can cover the commonly cited approximate human audible range, though practical filters need a transition band.
This is why CD audio uses 44.1 kHz.
What is Bit Depth?
Bit depth describes how precisely each sample is measured.
Bit depth is expressed in bits per PCM sample and primarily affects quantisation resolution and the potential noise floor. It is connected to available dynamic range rather than the highest audible frequency.
Common bit depths include:
Sixteen-bit PCM is used for conventional audio CDs; 24-bit recording is widely used during production. Twenty-four-bit input provides plenty of level-setting latitude even though real analogue hardware has a higher noise floor than ideal 24-bit quantisation.
How Bit Depth Works
Each sample is assigned a numerical value.
Higher bit depth means:
More bits provide more possible quantisation levels, reducing the error introduced by rounding sample values. They do not automatically make microphones, instruments or rooms quieter.
Dynamic Range
Bit depth directly affects dynamic range.
For ideal PCM quantisation, each additional bit improves theoretical signal-to-noise ratio by roughly 6 dB. A useful approximation is about 98 dB for 16-bit and 146 dB for 24-bit, before real-world analogue noise and converter limits.
Higher dynamic range allows for:
The low quantisation-noise floor available with 24-bit recording is useful when capturing quiet details and leaving sensible headroom for unexpected peaks. It does not remove acoustic noise or replace good microphone technique.
Sample Rate vs Bit Depth
These two concepts serve different purposes:
Sample rate controls the time spacing of measurements and the available frequency band. Bit depth controls the quantisation of each measured amplitude. They are not interchangeable forms of resolution.
Both are important for accurate digital audio.
In Music Production
Typical settings in a DAW:
For many projects, 44.1 or 48 kHz at 24-bit PCM is a practical setting. The appropriate choice depends on video requirements, delivery specifications and whether particular processing stages benefit from a higher sample rate.
Higher settings may be used, but they increase file size and processing requirements.
Why This Matters
Understanding sample rate and bit depth helps you:
Understanding the two controls helps you choose sensible recording settings, estimate storage requirements and recognise when a larger file offers no practical improvement for the intended output.
Sampling and reconstruction
At 48 kHz, an audio converter makes 48,000 measurements per second for each channel. The stored numbers are reconstructed into a continuously varying output through digital-to-analogue conversion and filtering. The smooth resulting signal should not be confused with a staircase diagram of the sample values.
The Nyquist limit and aliasing
The highest representable frequency in the ideal sampled system is below half the sample rate: 22.05 kHz at 44.1 kHz and 24 kHz at 48 kHz. Frequencies above the allowed band can fold into false lower components, called aliasing, unless appropriately filtered. Real converters also need a transition band for their filters.
Bit depth and quantisation noise
Linear PCM with 16 bits has 65,536 possible sample values, while 24 bits offers over 16 million. The familiar ideal sine-wave signal-to-quantisation-noise ratio is approximately 6.02N + 1.76 dB: about 98 dB at 16 bits and 146 dB at 24 bits. Hardware noise generally reduces the achievable real-world range.
Why dither matters
Quantisation introduces rounding error. When reducing bit depth, properly applied dither replaces potentially correlated low-level distortion with a very quiet noise floor. It is especially relevant when delivering a final 16-bit PCM file from a higher-precision mix, and should not be added repeatedly during routine editing.
Choosing project settings
For most music recording, 24-bit PCM at either 44.1 or 48 kHz is a sensible starting point. Video projects commonly use 48 kHz, while higher rates may be warranted for specific processing or delivery requirements. Uncompressed stereo 24-bit 48 kHz PCM requires about 17.3 MB of raw sample data per minute, excluding headers. Higher settings also increase storage and processing demands.
Related Topics
Related foundations include Digital vs Analogue Audio, The Decibel Explained and What Is Sound?.
This article is part of the Music Production Reference Library, organised into volumes covering sound, theory, production, and release.