.AAC - MPEG-2/MPEG-4 Advanced Audio Coding
ADTS/ADIF/RAW container formats
- Container formats: AAC only specifies an audio compression algorithm. The compressed data needs to be stored 'inside' a file format, referred to as a 'container format'. There are several such container formats: When it comes to formats using an .AAC file extension, then currently only the so called 'ADTS' format is supported for writing, while both 'ADTS' and 'ADIF' formats are supported for reading. The 'LOAS' and 'LATM' formats are not supported, nor is Real Audio (.RA files). The 'QuickTime' (.MOV files), 'MPEG-4 file format' (.MP4 and .M4A files), and 3GPP (.3GP and .3G2 files) container formats for AAC audio are also supported (both reading and writing) but are described separately.
- MPEG versions: There are both an MPEG-2 AAC format and an MPEG-4 AAC format. The latter has a few improvements in audio quality but may not be compatible with as much hardware and software. You can select which one you want when writing files.
- Profiles: There are various 'sub-formats' of AAC audio - called 'Profiles' - with varying levels of complexity. FAAC (see below) can write 'Main' and 'Low Complexity' profile files. The latter is compatible with the most hardware & software, while the former gives a bit better compression. FAAD 2 can read 'Main' (MAIN), 'Low Complexity' (LC), 'Scalable Sampling Rate' (SSR), 'Long Term Prediction' (LTP), 'High Efficiency' (HE), and 'Low Delay' (LD) profiles, including 'Error Resilient' (ER) variants.
- Bit-rate formats: When writing, you'll have two different types of selections to choose from in the data-formats list, 'CBR' and 'VBR':
Note that some AAC decoders only support 'CBR' files and the 'Low Complexity' profile.
- CBR (Constant Bit Rate) n kbit/s: This lets you (indirectly) specify the total size of the output file (file size = audio length * bit rate). The encoder then 'portions out' the total amount of available bits in variable amounts to different parts of the audio clip in such a manner as it thinks will produce the best overall audio quality.
- VBR (Variable Bit Rate) Quality n: This lets you specify a 'quality level' and the encoder then varies the bit rate so as to always achieve at least that given quality, without any concerns about what the total file size may end up being. This is the optimal way to compress the audio!
- Meta info: It's not an official part of any of the various AAC specifications, but meta-info structures are often added to .AAC files. Here's how that is handled:
- 'ID3 v1 meta info tags', 'ID2 v2 meta info tags' and 'APE meta info tags' are all read if present.
- Awave Audio only: 'ID3 v1 meta info tags' and/or 'APE meta info tags' can be written - if you enable it on the 'More options' page (click on the thus labeled button on the output options page). Note that this is turned off by default.
- 32-bit floating point data, up to 64 channels, 7 / 8 / 11 / 12 / 16 / 22 / 24 / 32 / 44 / 48 / 64 / 88 / 96 kHz, compresses to 8 kbit/s and up (minimum rate depending on channel count and sample rate), meta info tags.