Sample Rate and Bit Depth Explained Simply

What Is Sample Rate?

Sample rate is the number of times per second an audio signal is measured. The unit is Hertz (Hz) or kilohertz (kHz). A sample rate of 44,100 Hz means the audio was measured 44,100 times per second. This number comes from the Nyquist theorem, which states that to accurately reproduce a sound of a given frequency, you must sample at least twice that frequency. Human hearing tops out at around 20,000 Hz, so 44,100 Hz gives adequate headroom. The CD standard of 44,100 Hz and the video/broadcast standard of 48,000 Hz are by far the most common. Telephony often uses much lower rates: 8,000 Hz captures voice frequencies adequately and keeps data rates very low. You will also encounter 16,000 Hz in voice assistants and speech recognition systems, and 22,050 Hz in older multimedia content.

What Happens When Sample Rate Is Wrong

If you play back a PCM file at the wrong sample rate, you will hear a pitch and speed error. A file recorded at 48,000 Hz but played back at 44,100 Hz will play too slowly and sound lower-pitched — about 8% slower. Conversely, playing a 44,100 Hz file at 48,000 Hz makes it play fast and high-pitched. This is the most common symptom reported by people encountering PCM files for the first time: the audio sounds like a slowed-down or sped-up version of what it should be. The fix is simply to match the playback sample rate to the recording sample rate.

If you know what the audio should sound like, you can even work the true rate out: if speech sounds about 8% too slow at 44,100 Hz, the file was almost certainly recorded at 48,000 Hz (48,000 ÷ 44,100 ≈ 1.09).

What Is Bit Depth?

Bit depth determines how precisely each audio sample is measured. Think of it as the resolution of the measurement. An 8-bit recording can represent 256 different values for each sample. A 16-bit recording can represent 65,536 values. A 24-bit recording can represent over 16 million values. More values means finer gradations between loud and quiet — what audio engineers call dynamic range. Each additional bit of depth adds about 6 dB of dynamic range. 8-bit audio has roughly 48 dB of dynamic range — enough for speech but noticeably noisy. 16-bit audio has 96 dB of dynamic range — transparent for virtually all consumer listening. 24-bit audio has 144 dB of dynamic range — more than any recording scenario requires, but useful during mixing because headroom prevents clipping.

Common Bit Depth Values and What They Sound Like

8-bit audio has an audible noise floor and a characteristic hollow, slightly crunchy quality. You might recognize it from early computer game sound effects or very old telephone recordings. It is perfectly adequate for voice alerts, simple beeps, and non-critical audio but not for music or detailed speech. 16-bit is the CD standard and sounds clean and detailed to all but the most critical listeners. It is the right choice for distribution in nearly all consumer contexts. 32-bit float (also written as float32) is a special case — it uses floating-point numbers rather than integers, which gives exceptional headroom during processing. Raw PCM from some recording systems uses float32 internally. It sounds identical to 24-bit at the same sample rate but is easier to process without clipping artifacts.

Choosing the Right Settings for Your Situation

For voice memos and speech recordings where file size matters: 16,000 Hz or 22,050 Hz sample rate, 16-bit depth, mono channel. This produces clear, intelligible voice at a very small file size. For general music listening and podcast production: 44,100 Hz, 16-bit, stereo. This is CD quality and the safe default for any consumer audio work. For professional recording and mixing: 48,000 Hz, 24-bit, stereo. The extra headroom from 24-bit prevents clipping during editing. For telephony and IoT voice captures: 8,000 Hz, 16-bit (or 8-bit for the most constrained systems), mono. G.711 telephone audio uses 8,000 Hz specifically. For video production: 48,000 Hz is the broadcast standard and aligns with video frame rates more cleanly than 44,100 Hz.

How These Settings Affect File Size

The size of a raw PCM file is simply: sample rate × bit depth ÷ 8 × channels × duration in seconds. One minute of stereo audio at 44,100 Hz, 16-bit equals 44,100 × 2 × 2 × 60 = 10,584,000 bytes — about 10 MB. Compare that to one minute at 8,000 Hz, 8-bit, mono: 8,000 × 1 × 1 × 60 = 480,000 bytes — under half a megabyte. This explains why IoT devices favour low sample rates and bit depths, and why uncompressed WAV files get large quickly at high quality settings.

Summary

Sample rate and bit depth are the two numbers you need to know to work with any PCM file. Get the sample rate wrong and the audio plays at the wrong speed. Get the bit depth wrong and it turns into loud static or near-silence with glitches. Our raw audio troubleshooting guide walks through each symptom in detail. When you open a PCM file in PCM Drop, you can experiment with these settings in real time — the audio plays immediately so you can hear whether the settings are correct and adjust until it sounds right.