Pick a sample rate at the start of a project and you are committing to file sizes, CPU load and, to a small degree, sound quality. Should you record at 48,000 Hz or 96,000 Hz? For most people the answer is 48 kHz, and this guide explains why, and when 96 kHz is worth it.
What sample rate actually means
Sample rate is how many times per second the audio is measured. At 48 kHz, that is 48,000 snapshots a second; at 96 kHz, twice as many. The Nyquist theorem says a digital system can capture frequencies up to half the sample rate. So 48 kHz covers up to 24 kHz and 96 kHz covers up to 48 kHz.
Human hearing tops out around 20 kHz for young listeners and lower for most adults. That means 48 kHz already captures everything you can hear, with a little margin for the anti-aliasing filter.
48 kHz vs 96 kHz at a glance
| Factor | 48 kHz | 96 kHz |
|---|---|---|
| Frequency range captured | Up to 24 kHz | Up to 48 kHz |
| File size | Baseline | About double |
| CPU load | Lower | Noticeably higher |
| Plugin latency | Standard | Can be lower per sample, but heavier processing |
| Video and film work | Standard rate | Needs conversion |
| Audible benefit for listeners | Full audible range | Rarely audible |
Where 96 kHz can help
- Sound design and pitch shifting. If you slow recordings down a lot, the extra ultrasonic content becomes audible and the sound stays detailed.
- Heavy processing. Non-linear plugins such as saturators and distortion can generate aliasing at lower rates. A higher rate pushes aliasing further out of the audible range, though many plugins now oversample internally.
- Low latency monitoring. At a fixed buffer size, a higher rate means less time per buffer, though your computer works harder to keep up.
- Archival masters. Some engineers keep high-resolution sources for future use.
Where 48 kHz is the smarter choice
- Most music production, vocals, guitars and beats.
- Anything going to video, since 48 kHz is the video standard.
- Laptops and modest computers where CPU headroom matters.
- Large sessions with many tracks, where doubling the file size adds up quickly.
What about 44.1 kHz?
44.1 kHz is the CD standard and works fine for music that will only be released as audio. If your project is going to streaming and video, 48 kHz avoids a rate conversion. Whatever you choose, stay consistent inside the project; converting back and forth adds needless steps.
Bit depth matters more for recording
Recording at 24-bit gives you far more headroom against noise than 16-bit, and it is the real quality win for tracking. It also means you do not have to hit levels hot. Our guide to gain staging explains how to set sensible levels once you are recording at 24-bit.
How to choose in practice
- Decide the final destination: music only, or video too.
- Check your computer. If the session already strains at 48 kHz, 96 kHz will make it worse.
- Choose 48 kHz unless you have a specific reason for higher.
- Set the rate in the audio interface driver and in your DAW before you record anything.
- Keep the same rate for every file in the project.
Testing it yourself
If you are curious, record the same source at both rates and export at the same loudness. Ask a friend to swap the files without telling you which is which. Most people find it very hard to hear a difference in a blind test, which is a strong sign the extra cost is not worth it for regular work.
The bigger improvements come from better room treatment, a good microphone and a sensible performance. If you are still building your setup, spend on those first. You can also check our free tools in the free VST collection to improve your mixes without spending anything.
Key takeaways
- 48 kHz captures everything humans can hear and is the practical default.
- 96 kHz doubles file size and CPU load for a benefit few listeners can hear.
- Use 96 kHz mainly for sound design, extreme time stretching or specific workflow needs.
- 24-bit depth and good levels matter more than a higher sample rate.
- Pick a rate at the start and stay consistent.
FAQ
Is 96 kHz better quality than 48 kHz?
It captures higher frequencies, but those are outside human hearing. In normal music production the audible difference is very small.
Does 96 kHz reduce latency?
At the same buffer size it does lower the time per buffer, but the higher CPU load can force you to raise the buffer, which cancels the gain.
Should I record at 44.1 kHz instead?
It is fine for audio-only releases. Choose 48 kHz if the work may go to video.
Can I change the sample rate mid-project?
You can convert, but it is best to decide first and stay consistent to avoid conversion artefacts and problems.
