Short answer
For most home studios, 48 kHz at 24-bit is enough. Recording at 96 kHz doubles file size, roughly doubles the processing load for many plugins, and gives little audible benefit for typical music and voice recording. It’s worth considering in specific cases: heavy time-stretching or pitch-shifting, sound design, or plugin-heavy work where your plugins don’t oversample.
What 96 kHz changes
96 kHz raises the frequency ceiling to 48 kHz. That’s well beyond the roughly 20 kHz limit of human hearing. Blind tests generally show little or no audible difference at normal playback, though the topic is debated. Many microphones are designed around the audible range, so they capture little above 20 kHz.
A higher sample rate doesn’t improve your microphone, your room or your performance. Those usually matter far more than the sample rate.
Where 96 kHz can help
- Heavy processing. Nonlinear effects such as saturation, distortion and compression can create harmonics above the Nyquist limit. Those harmonics fold back as aliasing, and a higher sample rate gives them more room. Plugins with oversampling reduce the same problem, so the benefit shrinks if your plugins oversample.
- Stretching and slowing. Slow a 96 kHz recording to half speed and content up to 24 kHz remains. The same sound recorded at 48 kHz keeps only up to 12 kHz. This helps sound designers. Most home recording doesn’t need it. For the stretching tools themselves, see how to time-stretch audio.
- Latency. At the same buffer size in samples, a higher rate means less time per buffer. 128 samples is about 1.33 ms at 96 kHz, versus about 2.67 ms at 48 kHz. Total round-trip latency includes other delays too, so the real-world gain is smaller. A smaller buffer usually does more; see audio latency and buffer size.
What it costs
- Storage. 24-bit mono runs about 17.3 MB per minute at 96 kHz, versus about 8.6 MB at 48 kHz.
- CPU. Plugins and the DAW process twice as many samples per second, and many plugins need noticeably more processing.
- Interface channels. ADAT expansion uses S/MUX at higher rates. An 8-channel ADAT connection at 44.1 or 48 kHz drops to 4 channels at 88.2 or 96 kHz. Some USB interfaces also reduce channel counts or turn off onboard DSP at higher rates. Check your model’s manual.
- Converter quality. Some budget interfaces handle 96 kHz less well, and clock jitter can become more of a problem at higher rates.
- Playback. Heavily processed 96 kHz projects may not play back in real time on a modest computer.
When 48 kHz is enough
- You’re recording voice, guitar, drums or most instruments for a song, podcast or video.
- Your plugins oversample internally.
- You want to keep CPU use and disk space manageable.
- You plan to deliver at 44.1 or 48 kHz, which covers most streaming and video work.
A practical default
Use 48 kHz / 24-bit as your default. Consider 96 kHz for a specific project when you are doing heavy time-stretching or sound design, your plugins don’t oversample, and your computer and interface handle it well.
If you do go higher, 88.2 kHz pairs naturally with 44.1 kHz delivery, and 96 kHz pairs with 48 kHz for video. For the sample rate basics, see sample rate vs bitrate. For the 44.1 vs 48 kHz choice, see is 44.1 kHz or 48kHz better for music.
Common misconceptions
- “Upsampling 48 kHz to 96 kHz adds detail.” It doesn’t add information.
- “96 kHz files sound better because they’re hi-res.” A label doesn’t prove the source was recorded at that rate.
- “A higher sample rate fixes a poor mic or room.” Mic, room and performance matter far more.