Sample Rate Reference
Compare sample rates and bit depths. Calculate exact file sizes for any audio format, duration, and channel count.
| Sample Rate | Bit Depth | Nyquist Freq | Dynamic Range | MB/min (stereo) | Use Case |
|---|---|---|---|---|---|
| 44.1 kHz | 16-bit | 22.05 kHz | 96 dB | 10.1 MB | CD Audio Standard |
| 44.1 kHz | 24-bit | 22.05 kHz | 144 dB | 15.1 MB | Streaming / Download |
| 48 kHz | 24-bit | 24 kHz | 144 dB | 16.5 MB | Video / Broadcast Video Std |
| 88.2 kHz | 24-bit | 44.1 kHz | 144 dB | 30.2 MB | Hi-Res Music |
| 96 kHz | 24-bit | 48 kHz | 144 dB | 32.9 MB | Studio Recording Pro Std |
| 192 kHz | 24-bit | 96 kHz | 144 dB | 65.8 MB | Mastering / Archival |
| 192 kHz | 32-bit | 96 kHz | 192 dB | 87.9 MB | Highest quality |
| Bit Depth | Values | Dynamic Range | Noise Floor | Notes |
|---|---|---|---|---|
| 8-bit | 256 | 48 dB | -48 dBFS | Retro/lo-fi effects, game audio |
| 16-bit | 65,536 | 96 dB | -96 dBFS | CD standard. Sufficient for most playback |
| 24-bit | 16.7M | 144 dB | -144 dBFS | Recording & production standard. Plenty of headroom |
| 32-bit float | ~4B | ~1,528 dB | -768 dBFS | Floating point: immune to clipping during processing |
Sample rate and bit depth are the two numbers at the heart of every audio file. They determine how accurately digital audio represents the original sound. But there's an enormous amount of misinformation about what those numbers mean in practice, and a lot of producers are wasting CPU and storage on settings that provide no audible benefit.
This guide cuts through the confusion. You'll understand what sample rate and bit depth actually are, why 44.1kHz at 24-bit is the right choice for almost every recording situation, and when to deviate from that standard.
What Sample Rate Actually Means
Digital audio works by taking a snapshot of the audio waveform at regular intervals. Sample rate is how many of those snapshots happen every second. At 44,100 Hz (44.1kHz), the system captures 44,100 individual amplitude values per second. At 96kHz, it captures 96,000.
More snapshots per second means more detail; specifically, it means the system can accurately represent higher frequencies. The highest frequency a digital system can reproduce is exactly half its sample rate. This is the Nyquist theorem, and it's the single most important concept in digital audio.
The Nyquist Theorem: Why 44.1kHz Works
Human hearing tops out at around 20,000 Hz (20kHz). To capture that frequency accurately, you need a sample rate of at least twice that: 40,000 Hz. 44.1kHz gives you a comfortable margin: you can accurately reproduce up to 22,050 Hz, well above the threshold of human hearing.
This is why 44.1kHz became the standard for audio CDs in the 1980s, and why it remains the standard for streaming today. It contains all the audio information a human listener can perceive.
The practical implication: recording at 96kHz or 192kHz does not give listeners a better-sounding final product at 44.1kHz, because the extra high-frequency information gets discarded at export. What higher sample rates do offer is processing headroom: when you time-stretch, pitch-shift, or apply heavy processing, having more data to work with reduces artifacts. For that reason, some engineers prefer to record at 96kHz and export at 44.1kHz.
If you're not doing significant pitch correction or time-stretching, 44.1kHz is enough. The difference between 44.1kHz and 96kHz is not audible on consumer playback systems, but the file size and CPU load are twice as high at 96kHz. Record at 44.1kHz unless you have a specific reason not to.
Bit Depth: Resolution vs. Dynamic Range
Bit depth controls how many distinct amplitude values the system can represent. More bits = more possible values = finer resolution = greater dynamic range.
16-bit audio has 65,536 possible amplitude values. That gives you about 96dB of dynamic range: the difference between the softest sound and the loudest the format can represent without clipping. This is the CD standard.
24-bit audio has 16,777,216 possible values, giving you about 144dB of dynamic range. This is the recording and production standard.
32-bit float is used internally by most modern DAWs for processing. It has virtually unlimited headroom: you can clip your recording and recover the signal in the mix. 32-bit float files are large and not useful for distribution, but they're invaluable for session backups.
The difference between 16-bit and 24-bit matters most during recording. When you record with 16-bit, you have to hit your levels very precisely: too low and you're recording in the noise floor, too high and you clip. With 24-bit, you have 48dB more headroom. You can record conservatively (leaving plenty of space below 0dBFS) and still have excellent signal quality.
Sample Rate & Bit Depth Comparison
| Setting | Dynamic Range | Max Frequency | Best For |
|---|---|---|---|
| 44.1kHz / 16-bit | 96dB | 22kHz | CD, streaming delivery |
| 44.1kHz / 24-bit | 144dB | 22kHz | Recording, mixing, most projects |
| 48kHz / 24-bit | 144dB | 24kHz | Film, video, podcast |
| 96kHz / 24-bit | 144dB | 48kHz | High-res audio, heavy processing |
| 192kHz / 24-bit | 144dB | 96kHz | Archival, acoustic research |
| 44.1kHz / 32-bit float | ~1,500dB+ | 22kHz | DAW internal processing, session backup |
Which Settings to Use and When
For recording music: 44.1kHz or 48kHz at 24-bit. Use 44.1kHz if your final deliverable is music; use 48kHz if the music will be synced to video (48kHz is the video standard and avoids sample rate conversion artifacts).
For film/video scoring: 48kHz at 24-bit. Match the video standard. Converting between 44.1kHz and 48kHz introduces subtle artifacts. Avoid the conversion by working at 48kHz from the start.
For intensive pitch/time processing: 96kHz at 24-bit during production, export at 44.1kHz. The extra data helps algorithms produce cleaner results. Downsample at the final export stage.
Never record at 16-bit unless you have a specific creative or compatibility reason. The extra headroom from 24-bit costs nothing in practice and removes the stress of nailing perfect levels at record time.
Mixing sample rates in a project causes pitch and tempo problems. If your project is at 44.1kHz and you import a 48kHz sample, it will play back at the wrong pitch unless your DAW resamples it. Always check that samples, loops, and recorded tracks match your project sample rate before you record.
Correct Export Settings by Platform
Every distribution platform has a preferred format. Sending the wrong sample rate causes quality degradation through their own conversion process. Send the right format and you control how it sounds.
| Platform | Sample Rate | Bit Depth | Notes |
|---|---|---|---|
| Spotify / Apple Music | 44.1kHz | 16-bit | Accepts 24-bit; outputs at platform-specific rate |
| YouTube | 48kHz | 16 or 24-bit | Video standard; match if delivering with video |
| Mastering engineer | Match session | 24-bit | Send at your project's native sample rate |
| Podcast / Voice | 44.1kHz or 48kHz | 16-bit | Most platforms encode to compressed formats anyway |
| Sample libraries | 44.1kHz | 24-bit | Match the distribution standard for compatibility |
Frequently Asked Questions
What sample rate should I record at?
48 kHz for anything that may end up with video, 44.1 kHz for music-only work. Higher rates buy headroom for extreme processing at a real cost in file size and CPU, and the audible benefit is small.
What does sample rate actually control?
The highest frequency that can be represented: half the sample rate, the Nyquist limit. 44.1 kHz captures up to 22.05 kHz, comfortably above human hearing.
Is 24-bit better than 16-bit?
For recording, yes: not for detail but for headroom. 24-bit gives around 144 dB of dynamic range against 16-bit’s 96, so you can record conservatively. Record at 24, deliver at 16.
Does converting 48k to 44.1k lose quality?
Slightly, and it depends on the converter: it is not a clean integer division, so the resampler interpolates. Do it once, at the end, with a good converter; never repeatedly.
How big will my files be?
Roughly rate × bit depth × channels ÷ 8 bytes per second. Stereo 48 kHz 24-bit is about 17 MB a minute, so an hour across eight channels adds up faster than most people expect.