What Is AAC Format?
AAC (Advanced Audio Coding) is a lossy audio compression format developed as the successor to MP3. Created collaboratively by Fraunhofer Society, AT&T Bell Labs, Dolby Laboratories, Sony, and Nokia, AAC was standardized as MPEG-4 Part 3 (ISO/IEC 14496-3) in 1997 and became the default audio format for Apple's iTunes (2003), iPhone, iPad, YouTube, and countless streaming services.
At the same bitrate, AAC consistently produces better audio quality than MP3 — approximately equivalent to MP3 at 30% higher bitrate. A 128 kbps AAC file typically sounds as good as a 192 kbps MP3.
AAC Architecture: How It Differs From MP3
While both MP3 and AAC use psychoacoustic masking, AAC's encoding pipeline is more sophisticated:
| Component | MP3 | AAC |
|---|---|---|
| Filter bank | Polyphase filterbank + MDCT | Pure MDCT (no polyphase stage) |
| Transform size | 576 spectral lines | 1024 spectral lines |
| Temporal noise shaping | No | Yes (TNS) |
| Prediction | No | Yes (backward adaptive + intra-frame) |
| Joint stereo | MS stereo + intensity | MS stereo + intensity (improved) |
| Entropy coding | Huffman | Huffman (improved tables) |
| Filterbank resolution | Lower | Higher (better frequency resolution) |
The larger 1024-point MDCT in AAC provides better frequency resolution, reducing tonal artifacts. Temporal Noise Shaping (TNS) adaptively shapes quantization noise in the time domain, reducing pre-echo on transient sounds — one of MP3's most audible weaknesses.
AAC Profiles
AAC defines multiple profiles for different use cases:
AAC-LC (Low Complexity)
The most common profile. Used in iTunes, Apple Music, YouTube, most streaming services. Provides excellent quality at 128–256 kbps for music. No additional complexity features.
HE-AAC (High Efficiency / AAC+)
Adds Spectral Band Replication (SBR) to AAC-LC. SBR reconstructs high-frequency content from the low-frequency signal using a mathematical model, allowing the encoder to use fewer bits for high frequencies.
- Better than AAC-LC below 64 kbps
- Used in streaming radio (DAB+), online radio, some streaming services at low bitrates
- Adds ~2 ms latency from SBR processing
HE-AAC v2 (AAC+ v2)
Adds Parametric Stereo (PS) to HE-AAC. PS encodes stereo as mono plus a set of spatial parameters (inter-channel level and phase differences), reconstructing stereo on playback.
- Excellent quality at 24–40 kbps
- Used for very low-bitrate streaming radio, digital terrestrial radio (DAB+)
- Stereo at 32 kbps with HE-AAC v2 sounds better than MP3 stereo at 64 kbps
AAC-LD / AAC-ELD (Low Delay)
Low-latency variants designed for real-time communication (VoIP, video conferencing). AAC-ELD (Enhanced Low Delay) achieves ~20 ms codec delay suitable for two-way communication.
- Used in FaceTime audio, Apple's proprietary voice calls
- Not suitable for music broadcasting (higher distortion)
USAC (Unified Speech and Audio Coding)
The latest AAC variant (MPEG-D, MPEG-H Audio). Handles both speech and music efficiently in a single codec, designed for 8–48 kbps scenarios. Used in some DAB+ radio and emerging streaming applications.
AAC Container Formats
AAC audio can be stored in several container formats:
| Container | Extension | Description |
|---|---|---|
| MPEG-4 Part 14 | .m4a |
Standard Apple/iTunes container for audio-only AAC |
| MPEG-4 Part 14 | .m4p |
DRM-protected AAC (FairPlay) — legacy iTunes purchases |
| MPEG-4 Part 14 | .m4b |
Audiobook format with chapter markers |
| MPEG-4 Part 14 | .m4r |
iPhone ringtone format |
| MPEG-4 Video | .mp4 |
Video container that also contains AAC audio tracks |
| ADTS | .aac |
Raw AAC bitstream with ADTS headers — no metadata support |
| LATM/LOAS | N/A | Low-overhead transport for broadcast/streaming |
M4A is the standard file format for AAC music. It uses the same ISO Base Media File Format (ISOBMFF) as MP4 video. Unlike the raw .aac format, M4A supports rich metadata (title, artist, album, lyrics, album art) via iTunes Metadata atoms.
AAC Metadata
M4A files store metadata in iTunes-style atoms within the moov box:
| Atom | Field |
|---|---|
©nam |
Song title |
©ART |
Artist |
©alb |
Album |
©day |
Year |
trkn |
Track number |
©gen |
Genre |
©lyr |
Lyrics |
covr |
Album artwork |
aART |
Album artist |
cpil |
Compilation flag |
tmpo |
BPM |
©too |
Encoder tool |
Quality Comparison: AAC vs MP3
At equal bitrates, AAC is consistently superior to MP3. The differences are most noticeable at lower bitrates:
| Bitrate | MP3 Quality | AAC-LC Quality |
|---|---|---|
| 64 kbps | Acceptable speech | Good speech, acceptable music |
| 96 kbps | Fair music | Good music |
| 128 kbps | Good music | Very good music |
| 192 kbps | Very good | Excellent (near transparent) |
| 256 kbps | Excellent | Transparent |
| 320 kbps | Transparent | Transparent |
"Transparent" means the compressed version is indistinguishable from the original in a blind ABX test by trained listeners.
AAC Encoders
Apple AAC (CoreAudio / AudioToolbox)
Apple's proprietary encoder, considered one of the best AAC encoders. Available on macOS and iOS. iTunes and Logic Pro use this encoder. The only way to access it on non-Apple platforms is via Wine (running Windows iTunes) or cloud encoding services.
Fraunhofer FDK AAC
Open-source (LGPL) implementation released by Fraunhofer in 2012. Very high quality, comparable to Apple AAC. Available in FFmpeg (with FDK AAC support compiled in), Android's MediaCodec. Generally preferred over FFmpeg's native AAC encoder.
FFmpeg's Native AAC
Good quality at 192+ kbps, acceptable at lower bitrates. Available everywhere FFmpeg is available. Slightly lower quality than FDK AAC or Apple AAC, especially at low bitrates.
Encoding Commands
# FFmpeg with FDK AAC (best quality when available)
ffmpeg -i input.wav -c:a libfdk_aac -vbr 5 output.m4a
# FFmpeg native AAC at 256 kbps
ffmpeg -i input.wav -c:a aac -b:a 256k output.m4a
# FFmpeg HE-AAC v2 (for very low bitrate, e.g. 32 kbps)
ffmpeg -i input.wav -c:a libfdk_aac -profile:a aac_he_v2 -b:a 32k output.m4a
# iTunes/macOS afconvert (Apple encoder)
afconvert -f m4af -d aac -b 256000 input.aiff output.m4a
AAC in Apple Ecosystem
Apple's adoption of AAC:
- iTunes 2003: AAC replaced MP3 as default purchase format (FairPlay DRM initially)
- iPhone: AAC is the primary audio format; Voice Memos record in M4A
- Apple Music: streams at 256 kbps AAC (HLS)
- FaceTime audio: Uses AAC-ELD for low-latency voice
- HomePod/AirPlay: ALAC (lossless) or AAC-LC for music
- App Store audio: AAC for sound effects and music in apps
AAC in Streaming and Broadcasting
| Platform | AAC bitrate |
|---|---|
| YouTube (standard) | 126 kbps AAC-LC |
| Apple Music | 256 kbps AAC-LC (HLS) |
| Spotify (Very High) | 320 kbps OGG Vorbis (desktop) / 256 kbps AAC (mobile) |
| Netflix | HE-AAC and Dolby Audio (AC-4) |
| DAB+ Radio | HE-AAC v2 at 32–80 kbps |
| Podcasts (Apple Podcasts) | Typically 64–192 kbps AAC or MP3 |
Should You Convert MP3 to AAC?
No, not for existing MP3 files. Transcoding MP3 → AAC re-encodes a lossy file to another lossy format, causing generation loss (additional artifacts) even if the resulting AAC sounds similar. The information removed by the original MP3 encoding is permanently gone.
Yes, for new content. When capturing new audio or encoding from lossless masters (WAV, FLAC, AIFF), use AAC instead of MP3. You'll get better quality at smaller file sizes.
Exception: compatibility. If you need AAC files for an Apple ecosystem workflow (CarPlay, HomePod, iPhone) and you only have MP3 source files, converting to M4A may provide workflow benefits even if the quality cannot improve.
Related conversions
Audio format pairs that come up most often: