Published by PitchChanger · Updated · How these guides are made
Hear the difference
Owned one-second controls: identical PCM, with copied or changed WAV container bytes.
Pause your workspace preview before comparing these clips.
These controls illustrate file identity, not an audible quality comparison.
How to use this workflow
Decide which kind of equality matters
For backup verification, ask whether every byte matches the original. For audio comparison, ask whether decoded samples, sample rate and channel layout match. For delivery, also check required metadata and naming. Two files can contain identical PCM but have different headers or tags; matching duration alone establishes neither kind of equality.
Check bytes before interpreting a difference
Use a local SHA-256 tool, such as shasum -a 256 on macOS/Linux or Get-FileHash -Algorithm SHA256 in PowerShell. Matching cryptographic hashes are strong evidence of identical bytes. If the hashes differ, inspect decoded PCM before declaring that the sound changed. The converter itself is not a hash comparison tool; those checks happen outside PitchChanger.
Actual owned WAV controls: changed header, unchanged samples
We generated a one-second, mono, 48 kHz, 16-bit PCM sine control and copied it byte for byte. A third WAV uses the same PCM with an extra valid RIFF JUNK chunk containing four zero bytes. We compared file hashes and PCM payload hashes locally. The source construction and hashes are saved with the demonstration; no performer recording or perceptual test is involved.
The exact copy has the same file hash. The JUNK-chunk version has a different file hash and is 12 bytes longer, while its decoded PCM bytes match. This isolates a container-only change. It does not show that any arbitrary pair of differently encoded music files sounds identical.
Swipe or scroll sideways to read all columns.
| Comparison | Whole-file bytes | PCM payload |
|---|---|---|
| Control A vs copied B | Identical, 96,044 bytes each | Identical, 48,000 frames |
| Control A vs header-changed C | Different: C is 96,056 bytes | Identical, 48,000 frames |
| Same duration only | Insufficient evidence | Insufficient evidence |
Use the right conclusion for a mismatch
A differing file hash means at least one byte differs; it does not identify whether the change is a tag, encoded data, padding or corruption. A PCM comparison must also retain sample rate and channel order, since the same values interpreted at a different rate or layout do not describe the same playback.
For lossy re-encoding, expect different decoded values. A sample subtraction or equality check is not a perceptual quality score. Use a controlled listening comparison for audible differences, and preserve the original when the requirement is an exact archival copy.
Common questions
Can identical filenames prove that two exports match?
No. A filename is a label. Compare the files themselves. Conversely, renamed copies can contain exactly the same bytes, as the A/B control demonstrates.