Bitrate, Sample Rate, and Bit Depth, Explained

Every audio file wears a few numbers like a badge: 320 kbps, 16-bit / 44.1 kHz, 24/96. Forums argue about them, stores advertise them, and players display them, and yet bitrate, sample rate, and bit depth are mixed up constantly, even by people selling audio gear. They measure three completely different things. Untangle them once and a lot of audio marketing quietly deflates.

Here's each one on its own, then how they fit together.

Sample rate: how often the audio is measured

Digital audio is a series of snapshots. Thousands of times per second, the recording equipment measures the audio waveform's level and writes the measurement down. Sample rate is how many of those snapshots happen per second, measured in hertz. A sample rate of 44.1 kHz means 44,100 measurements every second, per channel.

What does that buy you? Frequency range. A foundational result in signal processing (the Nyquist theorem) says a digital system can perfectly capture any frequency up to half its sample rate. At 44.1 kHz, that's everything up to about 22 kHz. Human hearing tops out around 20 kHz in the young and lucky, and the ceiling drops with age.

So 44.1 kHz already covers the entire audible spectrum, with margin. Higher rates like 96 or 192 kHz extend the range further into ultrasonic territory, genuinely useful in production, where processing benefits from the headroom, but a much shakier proposition as a playback upgrade. That argument gets its own treatment in the hi-res audio myth.

One image to delete from your head: samples are not stair-steps. The reconstruction process on playback doesn't connect the dots with jagged edges: it recovers a smooth wave, and within the frequency limit, the original wave.

Bit depth: how precisely each measurement is stored

Each of those snapshots has to be written down as a number, and bit depth is how many bits that number gets. With 16 bits there are 65,536 possible values per sample; with 24 bits, about 16.7 million.

Here's the part almost everyone gets wrong: bit depth is not "detail" or "sharpness." What it actually sets is the noise floor, how far below the loudest possible sound the system's own background hiss sits. Each bit adds roughly 6 dB of dynamic range, so 16-bit gives you about 96 dB and 24-bit a theoretical 144 dB, more than any playback chain or listening room can use.

For scale: 96 dB is roughly the gap between a quiet room and a sound loud enough to hurt. With dithering (a standard technique that trades quantization error for a whisper of benign noise) 16-bit playback is effectively transparent. 24-bit earns its keep in recording and mixing, where levels are conservative and processing stacks up, which is why studios use it and why it's a sensible capture format. As a playback format, it's headroom you'll never hear into.

Bitrate: how much data flows per second

Bitrate is the simplest of the three: how many bits of data the file spends per second of audio, usually in kilobits per second (kbps). For uncompressed audio it's pure arithmetic:

44,100 samples × 16 bits × 2 channels = 1,411,200 bits per second, about 1,411 kbps for uncompressed CD audio.

But bitrate means two very different things depending on the kind of file you're looking at, and this is the confusion that launches a thousand forum threads.

For lossy files, bitrate is a budget

An MP3 or AAC encoder is handed a target (128 kbps, 256, 320) and must throw away enough audio to fit inside it. The bitrate is a choice, and it's a quality decision: the lower the budget, the more the encoder discards. That's why 320 kbps MP3s sound better than 128 kbps ones, and why the FLAC vs MP3 comparison is really a conversation about what gets deleted.

For lossless files, bitrate is an outcome

A FLAC file doesn't have a bitrate setting in any meaningful sense. It compresses the audio as far as it can without losing anything, and the resulting bitrate simply reports how compressible the music was. Sparse solo piano squeezes down a long way; dense, loud, noisy music doesn't. Two FLAC files (one at 600 kbps, one at 1,000 kbps) are both bit-perfect copies of their sources. Neither is "higher quality." One just contains music that's harder to squash. What is FLAC covers why nothing is lost either way.

For MP3, bitrate is a choice that costs you audio. For FLAC, it's just a report of how hard the file was to squeeze.

What "CD quality" actually means

Put the first two numbers together and you get the most famous spec in audio: 16-bit / 44.1 kHz, the CD standard, fixed in the Red Book specification around 1980. The 16 bits deliver ~96 dB of dynamic range; the 44.1 kHz covers the audible band. (Why 44.1 exactly, rather than a rounder number? Early digital audio was stored on adapted video equipment, and 44.1 kHz fit its geometry. The odd number is a historical fingerprint, not an acoustic one.)

Two things follow. First, "CD quality" describes a resolution, not a format: a FLAC, WAV, or ALAC file at 16/44.1 all qualify, because they carry the samples intact. A 320 kbps MP3 does not, however good it sounds, because samples have been irreversibly approximated. Second, CD quality is not the floor of respectable audio; by any honest measure of human hearing, it's at or beyond the ceiling.

Where bitrate, sample rate, and bit depth get confused

• "Higher sample rate = more detail I can hear." It captures higher frequencies, ones above your hearing range. Audible-band detail is already fully covered at 44.1 kHz.
• "24-bit sounds sharper." Bit depth sets the noise floor, nothing else. 16-bit's floor is already below audibility in playback.
• "This FLAC is 1,000 kbps and that one's 700, so the first is better quality." Lossless bitrate is an outcome of the music, not a quality dial. Both are perfect copies.
• "Bitrate and bit depth are the same thing." Bit depth is per sample; bitrate is per second. Related by multiplication, not by meaning.
• "Digital audio is stair-steps, analog is smooth." Reconstruction produces a smooth waveform. The stair-step diagram is a teaching sketch, not what comes out of your DAC.

The honest takeaway: these numbers describe how faithfully a recording was captured and stored: they say nothing about whether it was recorded or mastered well. A beautifully mastered 16/44.1 album will beat a brickwalled 24/192 release every single time, and no spec on the box will warn you.

If you want to see the numbers in the wild, open the player and drop in a file: it reads out the codec, sample rate, bit depth, and live bitrate, and with a FLAC you can literally watch the bitrate ride up and down with the music. The whole article, demonstrated in one readout.