Bitrate, Sample Rate and Bit Depth Explained

How digital audio works, in one paragraph

A microphone turns sound into a continuously changing voltage. To store it digitally, a converter measures that voltage thousands of times per second; each measurement is a sample. How often it measures is the sample rate, and how precisely it records each measurement is the bit depth. Multiply them together (and by the number of channels) and you get the bitrate of uncompressed audio. Lossy formats like MP3 then shrink that bitrate by discarding detail.

Sample rate (kHz)

The sample rate is the number of samples taken per second. According to the Nyquist–Shannon sampling theorem, a sample rate can capture frequencies up to half its value. Human hearing tops out around 20 kHz (lower as we age), so:

For listening, 44.1 or 48 kHz is all you need. Lowering the sample rate is a useful trick for speech: voices carry most of their intelligibility below 8 kHz, so a 22 kHz recording sounds fine and leaves more bits for what matters at low bitrates.

Bit depth (bits)

Bit depth is how many binary digits describe each sample. More bits mean finer steps between quiet and loud, which lowers the background noise floor. Each bit adds roughly 6 dB of dynamic range:

• 16-bit: about 96 dB of dynamic range. That's the CD standard, and more than enough for finished music.
• 24-bit: about 144 dB in theory. Used for recording, because it lets you record at safe, conservative levels without the quiet parts sinking into noise.
• 32-bit float: can represent levels above 0 dBFS without clipping, which is handy as an intermediate format while mixing and processing.

Bit depth only applies to uncompressed and lossless formats such as WAV and FLAC. MP3 and AAC don't store a fixed bit depth.

Bitrate (kbps)

Bitrate is the amount of data used per second of audio, in kilobits per second. For uncompressed audio it follows directly from the other two numbers:

CD-quality WAV: 44,100 samples × 16 bits × 2 channels = 1,411,200 bits per second ≈ 1,411 kbps.

Lossy encoders let you choose the bitrate. Lower bitrates mean smaller files and more aggressive removal of detail:

Constant vs variable bitrate

CBR (constant bitrate) uses the same data rate throughout, so file sizes are perfectly predictable. VBR (variable bitrate) spends more bits on complex passages and fewer on simple ones, often giving better quality per megabyte. Both are fully compatible with modern players. The tools on this site encode CBR, so the size estimates you see are exact.

How to calculate file size

For any format with a known bitrate:

Size (MB) ≈ bitrate (kbps) × duration (seconds) ÷ 8,000

For example, a 45-minute lecture at 64 kbps is 64 × 2,700 ÷ 8,000 ≈ 21.6 MB. The same lecture as CD-quality WAV would be about 476 MB. The Audio Compressor shows this estimate live as you change settings.

Recommended settings

Frequently asked questions