Sample Rates and Bit Depth, Straight Up

If you've spent any time scrolling through audio forums, you've seen the arguments. 44.1 versus 48. 16-bit versus 24. Some claim you need 192 kHz to capture the "air" in a string section; others say anything above CD quality is a scam. Both camps are usually talking past the actual point. Here's what matters when you're scoring a game, cutting a trailer, or prepping a playlist for release.

Let's start with sample rate. It's simply how many snapshots of a waveform we take each second. Compact discs run at 44,100 samples per second. Most video and game audio pipelines standardize on 48,000. The Nyquist theorem says you can perfectly reconstruct any frequency up to half the sample rate, so 44.1 kHz covers everything up to 22.05 kHz — well past the roughly 20 kHz limit of human hearing. Going higher doesn't give you more audible detail; it just moves the ceiling further away. That can matter for processing, not for listening. If you're doing heavy pitch-shifting or time-stretching, a higher rate gives the algorithm more data to work with. But for a finished orchestral stem that's just being mixed? 48 kHz is plenty. Don't let anyone sell you 192 kHz as a magic bullet.

Bit depth is a different animal. It's the vertical resolution — how precisely each sample's amplitude is measured. 16-bit gives you 96 dB of dynamic range. 24-bit gives you 144 dB. In practical terms, that extra headroom means you can record quieter and still have a clean signal when you normalize later. For game scoring, where you might deliver stems that get re-balanced by an audio director, 24-bit is the safe choice. It also helps when you're stacking dozens of layers — the noise floor stays put instead of creeping up with every gain stage. The difference isn't audible in a single track played back at normal levels, but it absolutely shows up in a complex session.

So where does that leave the working composer or curator? For delivery, match the destination. A Spotify release is going to end up at 44.1 kHz, 16-bit after encoding anyway. A game engine like Wwise or FMOD typically prefers 48 kHz, 24-bit source files. A film dub stage might ask for 48 kHz, 24-bit or even 96 kHz if they're doing extensive surround processing. The mistake is treating any one setting as universally "best." It isn't. It's just a tool for a job.

When you're recording, the rule is simpler: capture at the highest rate and depth your system can handle comfortably. Storage is cheap. CPU isn't infinite, but a 96 kHz session with 20 tracks is still manageable on a modern laptop. The reason to record high and mix low is that you can always downsample later — you can't add information back. That said, don't record at 192 kHz just because you can. Some plugins behave worse at extreme rates, and the file sizes get absurd. 48 kHz or 96 kHz at 24-bit covers 99% of professional work.

For playlist curators, the takeaway is different but related. You're often dealing with files from a dozen different sources — some mastered at 44.1, some at 48, some at 24-bit, some at 16. Your job is to make them sit together. That means checking for sample-rate mismatches before you sequence. A 48 kHz track dropped into a 44.1 kHz playlist will either get resampled by your player (usually fine) or play back at the wrong pitch (not fine). Most DAWs and editing tools handle this automatically, but it's worth a quick look. Bit depth matters less for playback — 16-bit is transparent for listening — but if you're doing any volume automation or EQ moves, working in 24-bit or 32-bit float internally avoids rounding errors.

None of this is glamorous. It's plumbing. But knowing why you're choosing a sample rate or bit depth — instead of just defaulting to whatever the preset says — makes you faster and harder to fool. And in a field where everyone has an opinion, that's worth more than a spec sheet.