Files arriving in a project rarely come from a single source. A field recorder, a camera, a sound library and an editing session may each use a different sample rate.
Most of the time, this is not something to change arbitrarily. Sample rate conversion becomes useful when a project, an application, a delivery specification or an established workflow requires a particular rate.
WHAT IS SAMPLE RATE?
Digital audio represents a sound as a series of measurements, or samples, taken at regular intervals. The sample rate is the number of samples taken per second, expressed in hertz.
The sample rate sets the upper limit of the frequencies a file can represent: roughly half the sample rate. Whether a recording actually contains content up to that limit depends on the microphones, the recorder and the source itself.
WHY CONVERT A SAMPLE RATE?
Conversion should answer a workflow requirement. Typical situations include matching the sample rate of a film or video project, preparing recordings for a post-production session, bringing together material recorded on different devices, meeting a delivery specification, or standardizing files within a sound library or editing workflow.
If none of these apply, the file can usually stay as it is.
44.1 KHZ, 48 KHZ AND 96 KHZ
44.1 kHz is commonly associated with music and audio distribution workflows.
48 kHz is a standard rate commonly encountered in film, video and production sound workflows. Location sound delivered to picture editorial is frequently expected at this rate.
96 kHz is often used for high-resolution recording. It can be useful in some field recording and sound design workflows, for example when material may later be pitched down or heavily processed.
None of these rates is universally superior. The right one depends on the recording, its intended use and the workflow it will enter.
DOWNSAMPLING
Downsampling produces a file with a lower sample rate. Because the new rate can represent a narrower frequency range, proper conversion filters out content above the new limit before resampling. Without that filtering, high-frequency content can fold back into the audible range as unwanted artifacts, known as aliasing.
Under normal sample rate conversion, the duration and pitch of the audio stay the same. Only the sampling grid changes.
UPSAMPLING
Upsampling produces a file encoded at a higher sample rate. The converter calculates new sample values between the existing ones, but it cannot recover frequencies or detail that were never captured in the 48 kHz source.
Upsampling changes the sampling grid of the file; it does not turn a lower-rate original recording into a higher-resolution original recording.
Upsampling is useful when a session or specification requires a higher rate, not as a way to improve an existing recording.
SAMPLE RATE VS BIT DEPTH
Sample rate and bit depth are often written together, as in 48 kHz / 24-bit or 96 kHz / 24-bit, but they describe two different things.
A 48 kHz / 24-bit file and a 96 kHz / 24-bit file share the same bit depth but differ in sample rate. Changing the bit depth is a separate operation and is not sample rate conversion.
WHEN SHOULD YOU CONVERT?
There is generally a reason to convert when the project requires another sample rate, when a delivery specification calls for it, when files need to match an established session or workflow, or when an application or device requires it for compatibility.
If the existing file already matches the required workflow, conversion is usually unnecessary.
CONVERT A SAMPLE RATE
If you need to prepare an audio file for a workflow that requires a different sample rate, a sample rate converter can resample the file to the required rate. One option is the Sample Rate Converter — Audio Tool Kit.