What is sample rate? It is the number of times per second an audio signal is measured during analogue-to-digital conversion. For voice and podcasts, 44.1 kHz or 48 kHz are both common choices; pick one and keep it consistent.
Settings only go so far with a weak source, so start with the best microphone for voice recording you can reasonably get, then use the rest of this guide to set up the session around it.
How sampling turns sound into numbers
Sample rate is the sampling frequency of a digital recording. Yamaha’s glossary defines it as the number of times per second that audio signals are sampled, meaning measured, during A/D conversion. Waveroom puts the same idea as the number of samples of audio taken per second. A microphone produces a continuous electrical signal, and the converter in your interface or recorder cannot store a continuous signal. It stores snapshots instead.
The process runs in stages, and each stage has its own setting:
- Measure at regular intervals. The converter reads the incoming signal over and over. The sample rate sets how many readings it takes each second.
- Store each reading as a number. The precision of each reading is the bit depth, which refers to the number of bits used to represent each sample.
- Keep the sequence. The resulting stream of numbers is the recording. A higher rate provides more data points per second.
Sample rate and bit depth are therefore separate controls. One decides how often the signal is read, the other decides how finely each reading is described. The next question is how often is often enough, and that is what the Nyquist theorem answers.
Nyquist limit and aliasing
Half the sample rate is the ceiling
The Nyquist theorem says that you need a sampling rate equal to twice the highest frequency of a signal. Turned around, the highest frequency that can be captured at any sample rate is half that rate. At 48 kHz sampling, the highest capturable frequency is 24 kHz. In practice, usable bandwidth is a bit less than half of the sample rate, so treat half the rate as a ceiling rather than a target.
This is why the common rates exist. Routenote notes that most human ears perceive sounds from 20 Hz to 20 kHz, and the usual rates keep that whole range below the ceiling.
What aliasing does to a recording
Aliasing happens when the input contains frequencies above half the sample rate. Those frequencies appear to be lower than they actually were. Ginger Audio’s example is a 25 kHz signal sampled at 48 kHz, which looks like a 23 kHz signal. Learning Modular describes the audible result as an additional set of higher harmonics that mirror the correct ones. The distortion is baked into the sampled data, so you should prevent it rather than plan to fix it afterwards.
Interfaces include anti-aliasing filters for this reason, but Ginger Audio cautions that the filters are not completely effective, so the frequency content of what you record still matters. For speech this seldom becomes a practical problem, because the voice carries little energy in the range where aliasing starts.
Common sample rates and what they suit
Yamaha lists 32 kHz, 44.1 kHz and 48 kHz as common digital audio sampling rates, and audio interfaces typically offer 44.1 kHz or 48 kHz. Higher rates sit above them for specialist work. The table below puts the usual options side by side.
| Sample rate | Samples per second | Planning use |
|---|---|---|
| 44.1 kHz | 44,100 | The most commonly used rate in music production and podcasting; a safe default for spoken-word work. |
| 48 kHz | 48,000 | A standard interface option; Waveroom records podcasts as uncompressed audio at this rate. |
| 88.2 kHz | 88,200 | More data points per second and larger files; a choice for music recording rather than speech. |
| 96 kHz | 96,000 | Same trade-off as 88.2 kHz, with the same larger files. |
| 192 kHz | 192,000 | Its necessity is debated among professionals; hard to justify for voice. |
Sample counts come from RouteNote’s sample rate guide, and the podcasting use is described by Waveroom.
The practical reading is simple. The two lowest rows in the table cover almost every spoken-word project. The higher rows buy extra data per second, and that has a cost in storage, which the next section explains.
Bit depth, bit rate and file size
File size depends on more than sample rate. Waveroom shows that bit rate is bit depth multiplied by sample rate: 16-bit audio at 44.1 kHz has a bit rate of 705,600 bits per second (16 x 44,100). Raise either number and the data rate rises with it. RouteNote states the consequence plainly: recording at a higher sample rate leads to a bigger audio file.
Ginger Audio notes that interface resolution is typically either 16 bits or 24 bits. Because both settings multiply the data, the sensible way to control file size is to avoid raising either one without a reason. A long interview recorded at a rate far above what the voice needs fills drives and slows backups with no matching gain in what listeners hear.
Does a higher sample rate improve quality?
Yes, in a limited sense. A higher rate allows for a broader frequency response and improved resolution, and it gives more data points per second. The same source says the necessity of ultra-high rates such as 192 kHz is debated among professionals. The cost, a bigger file, is not debated.
For speech the balance is clear. The rates in the table already place the ceiling above the commonly cited 20 kHz limit of hearing, so extra headroom mostly adds storage rather than audible clarity. Audio quality in voice work is decided far more by the room and the microphone. Directional pickup is one example, covered in Cardioid Polar Pattern: The Essential Guide To Directional Miking In 2026 – Voiceoriginals.
Recommended sample rate for voice and podcasts
Record voice and podcasts at 44.1 kHz or 48 kHz, and prefer 24-bit. Waveroom reports that 44.1 kHz is the most commonly used rate in music production and podcasting, which makes it the lowest-friction default. The same source describes 24-bit as currently considered the gold standard, which is a good reason to choose it over 16-bit when you plan to edit.
Choose between the two rates like this:
- Choose 44.1 kHz if you want the most common podcast setting and the smallest files of the two.
- Choose 48 kHz if your recording platform or interface works natively at that rate. Waveroom itself records podcasts as uncompressed audio at 48 kHz.
- Skip the ultra-high rates for speech unless a client specifies them. You pay in file size for a benefit that, at the top end, professionals still debate.
Then set the same rate in both places where it lives: the audio interface or recorder and the project in your editing software. Set the rate before you hit record, so the project matches what the converter delivers.
If you are still choosing hardware, the interface connection matters as much as the setting. What Is An XLR Microphone? The Essential 2026 Guide To Pro Audio Connections – Voiceoriginals explains how professional microphones plug in.
Frequently asked questions
Is sample rate the same as bit rate?
No. Sample rate counts measurements per second, while bit rate is the bit depth multiplied by the sample rate, so it also reflects how many bits describe each measurement.
Which sample rates do audio interfaces usually offer?
Ginger Audio describes the common rates as 44.1 kHz or 48 kHz. Yamaha also lists 32 kHz among the common digital audio sampling rates.
Is 16-bit still acceptable for voice?
It is a recognised standard, because 16-bit resolution is the same as used on CDs. Recording at 24-bit is the better choice when you intend to edit, as covered above.
