Sample Rate and Bit Depth: What They Do in Recording

Sample rate and bit depth describe how a digital recording stores sound. Sample rate sets how often the signal is measured each second, which determines the highest frequency the recording can hold. Bit depth sets how finely each measurement is stored, which determines the range between the quietest and loudest sounds. Both settings are chosen when you set up a project, and they affect file size, processing load, and sound quality.

What digital audio is

An analog signal is a continuous voltage that follows the sound pressure wave. A converter in the audio interface turns it into digital numbers. It takes a series of snapshots of the voltage at regular intervals, and each snapshot is rounded to the nearest value that the bit depth allows. The series of numbers is the recording. Playback reverses the process, turning the numbers back into a voltage that drives the speakers.

Sample rate and the frequency limit

The sample rate is the number of snapshots per second, measured in hertz or kilohertz. The Nyquist-Shannon sampling theorem says that a sampled signal can represent frequencies up to half the sample rate. This upper limit is called the Nyquist frequency. At 44.1 kHz, the limit is 22.05 kHz. At 48 kHz, it is 24 kHz. At 96 kHz, it is 48 kHz.

Human hearing typically extends to about 20 kHz, though the upper limit falls with age and exposure. A sample rate above 40 kHz is therefore enough to capture the full audible range. This is why 44.1 kHz was chosen for CD audio, and why 48 kHz became the standard for video and broadcast.

Why higher sample rates exist

If 44.1 kHz covers the audible range, why do higher rates exist? Part of the answer is headroom for processing. Many plugins, especially distortion and other nonlinear effects, create harmonics above the original frequencies. At a higher sample rate, these harmonics have more room before they fold back into the audible band as aliasing. Oversampling inside plugins addresses the same problem, which is one reason some engineers record at 96 kHz.

The cost is real. Doubling the sample rate doubles the data each second, increases file sizes, and raises the CPU load. The benefit is smaller than marketing often suggests for ordinary material. A well-designed converter at 44.1 or 48 kHz sounds excellent, and differences at 96 kHz are hard to hear in controlled listening tests of typical program material.

Bit depth and dynamic range

Bit depth is the number of bits used to store each snapshot. Each extra bit doubles the number of possible values. For linear PCM, the theoretical dynamic range is about 6 dB per bit, plus a small constant. A 16-bit file therefore offers about 98 dB of theoretical range, and a 24-bit file offers about 146 dB.

These are theoretical limits. A real converter has noise from its electronics, and the room has noise too, so the usable range is smaller. Even so, the difference is large. A 16-bit recording is limited by the CD standard, while 24-bit gives recording engineers much more room to set levels without the noise floor intruding.

Floating point inside the DAW

Most DAWs process audio internally in 32-bit or 64-bit floating point. Floating point stores a mantissa and an exponent, so the precision stays roughly constant across a wide range of levels. This means a mix can exceed full scale temporarily without clipping inside the DAW, as long as the level is brought back under full scale before the output. Fixed-point file formats used on export are where clipping matters.

Comparing common settings

The table below lists the usual combinations and where each one fits. These are common practical choices rather than rules.

Format	Sample rate	Bit depth	Typical use
CD audio	44.1 kHz	16-bit	Final music distribution
Video and broadcast	48 kHz	24-bit	Film, TV, and online video
Studio recording	48 kHz or 96 kHz	24-bit	Tracking and mixing
Internal processing	Project rate	32-bit float	Inside the DAW

File size and processing load

File size scales with both settings. A mono 24-bit file at 48 kHz uses about 144 kilobytes per second, which is roughly 8.6 megabytes per minute. Doubling the sample rate or adding channels increases the cost in proportion. A multitrack session with dozens of takes adds up quickly, so plan storage accordingly. Processing load follows the same logic, because plugins run once per sample, and higher rates mean more work per second.

Which settings to use

For most home projects, choose 48 kHz and 24-bit. This matches common video and broadcast standards, leaves comfortable headroom, and keeps files manageable. If you deliver to a video project, match its sample rate. Use 44.1 kHz when the destination requires it, such as a CD master. Consider 96 kHz when you need extra headroom for heavy plugin processing and have the storage and CPU to spare.

Set the rate when the project is created rather than changing it later. Changing the rate mid-project can cause pitch and speed problems if the DAW does not resample correctly. Check the interface's supported rates before you start.

Dithering when reducing bit depth

When you export a 24-bit mix to 16-bit, the DAW must reduce the number of bits. Simple rounding adds distortion that is correlated with the signal. Dither adds a small amount of controlled noise before rounding, which decorrelates the error and keeps quiet detail audible. Most DAWs offer dither as an export option. Apply it once, at the final stage, and never stack it on intermediate bounces.

Resampling and conversion quality

Every time audio changes sample rate, a resampler recalculates the samples, and the quality of that calculation matters. A good resampler applies a steep low-pass filter just below the new Nyquist frequency, which prevents frequencies that cannot be represented from folding back as aliasing. Poor resamplers can introduce audible artifacts, especially on high-frequency material such as cymbals and sibilance. Most modern DAWs and operating systems include high-quality resampling, but cheap converters and older software may not.

The practical rule is to avoid unnecessary conversions. Import files at the project rate when possible, and export once at the final rate. Converting back and forth between 44.1 and 48 kHz adds small losses each time. When a client delivers material at a different rate, convert it once using the DAW's highest quality setting, and keep the original file as a reference.

Key takeaways

Sample rate determines the highest frequency a recording can contain, and half the rate sets the limit. Bit depth determines the usable dynamic range, with about 6 dB per bit in the theoretical ideal. Choose 48 kHz and 24-bit for most projects, and match the destination when delivering finished work.

Higher settings cost storage and processing power, and their benefit depends on the material and the processing chain. Use dither when reducing bit depth, and set the project rate before recording begins.