WAV vs MP3 for Sampling: What Actually Changes
Published July 21, 2026
"Always use WAV" is the advice you'll see repeated everywhere, and it's not wrong — but it skips the actual reason, which means people apply it in situations where it doesn't matter and ignore it in situations where it really does.
What's technically different
WAV stores audio uncompressed — every sample, bit for bit. MP3 throws away data the format's psychoacoustic model predicts you won't notice, which is how it shrinks a file to roughly a tenth the size. At 320kbps, that prediction is good enough that most listeners can't reliably tell the difference in a blind A/B test on a finished track.
The problem isn't how MP3 sounds on first listen — it's what happens when you process it further.
Where compression artifacts actually show up
MP3's data loss is concentrated in specific places: high-frequency detail, transient sharpness, and stereo image precision. On a finished, mastered track played straight through, that loss is usually inaudible. But sampling isn't playing a track straight through — it's pitching, stretching, EQ'ing, and layering small slices of it, which is exactly the kind of processing that exposes what's missing.
- Pitching a sample down stretches artifacts that were inaudible at original pitch into a range you can actually hear.
- Heavy EQ or saturation on a chopped sample amplifies whatever noise floor and frequency gaps the compression left behind.
- Layering multiple MP3 sources stacks artifacts that were each individually inaudible into something that reads as mud or harshness.
A practical way to decide
Ask what happens to the file after you get it, not what it's for in general:
- Chopping, pitching, or layering it into a beat → WAV. You're about to stress-test exactly the parts of the signal MP3 discards.
- Listening for reference, checking an arrangement, or studying a mix → MP3 is fine. You're not processing it further, just listening.
- Not sure yet what you'll do with it → grab WAV. Converting WAV down to MP3 later costs you nothing; going the other direction can't recover data that's already gone.
The file-size tradeoff is smaller than it seems
A three-minute WAV runs roughly 30MB versus 3MB for the same track as 320kbps MP3. That difference mattered when storage was expensive and slow; on current hardware, a full sample library in WAV is rarely the bottleneck it used to be. Unless you're archiving thousands of reference tracks, the storage math isn't a strong reason to default to MP3 anymore.