Audio Transcoding vs. Converting: What Is the Difference?

"Converting" and "transcoding" get used interchangeably in casual audio talk, but they describe different operations with different quality implications. Knowing which one you are about to perform tells you immediately whether the result will be bit-perfect, lossy-with-no-additional-loss, or genuinely degraded compared to your source. This is the explainer that draws the line clearly.

Converting: The Umbrella Term

Converting is the general term for moving audio from one format to another. It covers every possible source-and-destination pair:

Lossless to lossless (WAV → FLAC, FLAC → WAV, WAV → AIFF) — bit-perfect; the same PCM samples in a different container or compression scheme.
Lossless to lossy (WAV → MP3, FLAC → AAC, WAV → Opus) — first-generation lossy encode from a clean source; quality determined by the codec and bitrate.
Lossy to lossless (MP3 → WAV, OGG → FLAC) — repackages already-lossy audio into a lossless container; no quality recovery, no further loss.
Lossy to lossy (MP3 → AAC, OGG → MP3, AAC → OGG) — this is transcoding specifically, and it stacks lossy generations.

In casual conversation "I need to convert my MP3 to WAV" is fine and correct. The technical distinction matters when you are deciding whether the operation will degrade quality.

Transcoding: The Specific Lossy-to-Lossy Case

Transcoding is the proper name for decoding one lossy format and re-encoding the result as another lossy format. The mechanics:

1. The lossy source file is decoded back to PCM. The decoder reconstructs samples from the compressed bitstream — which already contains psychoacoustic artifacts baked in by the original encoder. 2. The reconstructed PCM is fed into a different lossy encoder. 3. The new encoder applies its own psychoacoustic model and discards data it judges inaudible.

Each lossy encode is irreversible. Stacking two of them stacks two generations of artifacts. Concrete examples:

MP3 → AAC: transcoding. The AAC encoder discards detail from MP3's already-imperfect output.
OGG → MP3: transcoding. Two different lossy schemes, two generations of loss.
YouTube AAC download → Opus for delivery: transcoding.
Old voice memo (Opus) → MP3 for sharing: transcoding.

Why Transcoding Degrades Quality

The image analogy is exact: save a photo as JPEG at 70% quality, open the JPEG, save again at 70%, and the second JPEG looks visibly worse than the first — not because JPEG quality dropped (you used the same setting), but because the encoder threw away different details on the second pass. Audio works identically.

Specific failure modes that compound through transcoding:

High-frequency cut. Most lossy codecs reduce or eliminate content above a certain frequency at low bitrates. Transcoding stacks two cuts.
Metallic / "watery" artifacts from psychoacoustic mismatch — the second encoder hears artifacts from the first encoder and tries to "preserve" them, often making them worse.
Sibilance and cymbal smearing become noticeably ratty after a single transcode at typical bitrates.
Stereo imaging collapse — joint stereo decisions from one encoder may interact poorly with the other's spatial coding.

A 192 kbps MP3 transcoded to 192 kbps AAC sounds worse than either 192 kbps MP3 from a clean source or 192 kbps AAC from a clean source. There is no recovery.

Lossless Conversion: No Quality Change at All

Conversions between lossless formats are not transcoding in the degrading sense. The PCM samples remain bit-identical:

WAV → FLAC: decode WAV's PCM, encode it as FLAC. Decoded FLAC is bit-identical to the original WAV. Verifiable with md5sum.
FLAC → WAV: reverse the above. Bit-identical PCM out.
WAV → AIFF: identical PCM in a different container.
FLAC → ALAC: decoded ALAC matches decoded FLAC exactly.

These conversions are completely safe. Repeat them as many times as you want.

Lossless to Lossy: First-Generation Lossy

Encoding WAV → MP3 or FLAC → Opus is a single lossy generation from a clean source. This is the right way to produce a lossy delivery file. The encoder has full original data to work with; it discards inaudible content once and writes the result. Quality depends on the codec and bitrate, but no compounding artifacts are introduced.

The rule: always encode lossy formats from a lossless source when possible.

Lossy to Lossless: Repackaging Without Recovery

Going MP3 → WAV or OGG → FLAC takes lossy audio and stores it in a lossless wrapper. The decoded PCM is preserved bit-perfectly going forward, but it cannot be better than the lossy source. Useful workflow when:

You will edit the audio in a DAW that handles WAV more cleanly than the source format.
You need to apply DSP and want every step from there to be lossless.
You're standardizing an archive.

Not useful when you expected quality improvement (it does not happen) or when storage matters (lossless wrappers are 5-10× larger).

Practical Decision Rules

Lossless source available? Encode directly from it. Never transcode through an intermediate lossy generation.
Need lossy output but only have lossy source? Pick the highest practical output bitrate. Match or exceed the source bitrate. Use the same codec if the destination supports it (re-saves with the same codec at the same settings can sometimes re-use frames; mostly still lossy but less degraded than cross-codec).
Editing audio? Decode to lossless once, edit there, encode to delivery format once at the end.
Just need playback compatibility? Lossy → lossy is fine for casual purposes; quality loss is small at decent bitrates and nobody is ABX-testing voice notes.

Why This Distinction Matters

The vocabulary matters because it predicts outcomes. "I'm converting WAV to FLAC" → bit-perfect, no concern. "I'm converting MP3 to AAC" → that's transcoding, expect quality loss. "I'm converting FLAC to MP3" → first-generation lossy from a clean source, expect a clean encode. The same word ("converting") covers all three but the actual operations behave very differently.

For more on the underlying concepts, see What is audio transcoding for the deeper codec-level mechanics, What is an audio codec for the encoder/decoder pair concept, and lossless vs lossy for the format-family decision tree. The audio bitrate guide covers how bitrate changes the perceived loss in each lossy generation.