--- title: "What MP3 and AAC Encoding Changes | Gawergy Audio" description: "Understand perceptual coding, irreversible information loss, possible artifacts and decoded peak changes, without treating every lossy file as unusable." canonical_url: "https://gawergy.com/learn/what-happens-when-audio-is-encoded" md_url: "https://gawergy.com/learn/what-happens-when-audio-is-encoded.md" last_updated: "2026-09-25" date_published: "2026-09-25" author: "Gawergy Audio" publisher: "Gawergy Audio" language: "en-US" article_section: "Mastering & Delivery" --- # What Happens When Audio Is Converted to MP3 or AAC? MP3 and AAC encoding transform a PCM source into a smaller lossy representation by allocating bits according to perceptual coding models. The decoded audio is a new approximation of the source, not an exact copy of its original samples. Quality depends on codec, implementation, settings, and material. A high-quality encode can be useful for listening or distribution, but keep a lossless master for revisions and future formats. ## Key takeaways - Lossy encoding discards or coarsens information; decoding does not recover the original PCM samples. - Encoder settings and source material influence audible artifacts more than a codec name alone. - A decoded file can have different sample and true peaks from the source. - Produce each lossy derivative from a lossless approved master when possible. ## A codec makes perceptual tradeoffs Fraunhofer's MP3 and AAC technical overview describes transform-based analysis, quantization, and perceptual models. The encoder tries to represent parts of the signal judged most important under a bit budget. This is not the same as FLAC, whose decoded PCM is bit-for-bit identical to the PCM it encoded. Both reduce file size, but only one is lossless. The decisions are implementation-specific. Two encoders using the same format and nominal data rate need not create identical output. A simple claim that AAC is always better than MP3 or that one bitrate is transparent for all music overstates the evidence. Listen to the actual derivative when it will be distributed widely or used as a client reference. ## What listeners and meters may notice Difficult material can reveal changed high-frequency texture, pre-echo around transients, or altered stereo detail, depending on codec settings. A carefully encoded file at a suitable rate may sound very close to its source in ordinary listening. The point is not to invent a universal audible threshold; it is to acknowledge that the waveform and available information have changed. A reconstructed waveform can also produce peaks between samples that were not apparent in the source's sample-peak reading. Apple Digital Masters emphasizes auditioning an AAC encode rather than assuming the PCM master alone proves the encoded result is safe. The [true peak after encoding article](https://gawergy.com/learn/why-true-peak-changes-after-encoding) explains that meter distinction. ## Why the lossless source matters If an MP3 is decoded and encoded again, the second encoder works on a signal already altered by the first. The second pass may compound or move artifacts. Converting the MP3 to WAV first does not recover discarded content; it merely stores the decoded approximation in a lossless container. A verified WAV or FLAC master avoids that unnecessary generation loss. Ableton Live 12 allows PCM export and a separate MP3 export. Apple documents tools to create and audition its AAC encoder output. Those are examples of deriving review formats from a controlled source. Keep the destination's specifications in view: an aggregator may request lossless PCM even if listeners eventually receive a lossy stream. ## Judge the derivative fairly Compare source and encode at similar playback level on the same section. Check quiet tails, bright percussion, distorted textures, and stereo effects if those are central to the song. Do not infer codec quality solely from the fact that two files fail a null test; lossy coding is expected to change the samples. The question is whether those changes harm the intended listening result. Keep track of what was encoded and at which settings. A later platform may transcode from a delivered master using its own pipeline, so your local MP3 is not necessarily the exact consumer stream. The [WAV/FLAC/MP3 comparison](https://gawergy.com/learn/wav-vs-flac-vs-mp3) helps choose the right source file for each stage. ## Sources & References - [Fraunhofer IIS: MP3 and AAC Explained](https://www.iis.fraunhofer.de/content/dam/iis/de/doc/ame/conference/AES-17-Conference_mp3-and-AAC-explained_AES17.pdf) - [Apple Digital Masters: Mastering for Apple Music](https://www.apple.com/vn/apple-music/apple-digital-masters/docs/apple-digital-masters.pdf) - [Ableton Live 12 Manual: Managing Files and Sets](https://www.ableton.com/en/manual/managing-files-and-sets/) - [Xiph.org FLAC: Features and lossless coding](https://www.xiph.org/flac/features.html) ## Continue Reading - [WAV vs FLAC vs MP3](https://gawergy.com/learn/wav-vs-flac-vs-mp3) - [Why True Peak Changes After Encoding](https://gawergy.com/learn/why-true-peak-changes-after-encoding) - [Does Exporting Twice Reduce Quality?](https://gawergy.com/learn/does-exporting-twice-reduce-quality) ## Sitemap See the full [Gawergy.com sitemap](https://gawergy.com/sitemap.md).