AAC-LC vs HE-AAC vs xHE-AAC: AAC Family Explained (2026)

Updated October 2026: xHE-AAC encoding is now required in Android 17, the current Via LA AAC royalty table, a corrected patent-pool history, and short answers to “AAC vs AAC-LC”, “HE-AAC vs AAC” and “what does AAC 2.0 mean”, plus an FAQ.

AAC-LC, HE-AAC and xHE-AAC: the short answer

AAC (Advanced Audio Coding) is a family of four audio codecs that all share one core. AAC-LC (Low Complexity) is that core and the default audio in MP4 video, streaming and phones; it sounds best from about 96 to 256 kbps stereo. HE-AAC v1 is AAC-LC plus Spectral Band Replication (SBR), which rebuilds the treble, for roughly 48–80 kbps. HE-AAC v2 adds Parametric Stereo on top, for roughly 24–48 kbps. xHE-AAC is the newest member, built on USAC, covering about 12 kbps mono speech up to transparent stereo with mandatory loudness metadata.

HE-AAC vs AAC-LC: HE-AAC wins only when bits are scarce. Below about 64 kbps stereo, HE-AAC sounds clearly better; above about 96 kbps, AAC-LC sounds as good or better and plays on more devices. If you can only remember one rule, it is this one: always name the member and profile, never just “AAC”.

Why this matters

If your product touches video — streaming, conferencing, OTT, telemedicine, e-learning — AAC is almost certainly already in your pipeline, because it is the default audio codec for MP4, YouTube, Netflix, Apple devices, and most of the web. But "we use AAC" is an incomplete sentence. The four family members behave so differently that confusing them is the audio equivalent of confusing a sedan with a freight truck. This article is written for a product manager, founder, or operations lead with no audio background; every claim traces back to the ISO/IEC standard or the patent maker's own statement, not a secondhand summary. Read it and you will know exactly which AAC to ask your engineers for, and why.

The family, in one sentence each

The word "AAC" stands for Advanced Audio Coding. It first appeared in 1997 as the successor to MP3, and it has grown into a family because engineers kept adding optional layers on top of the original codec rather than replacing it. Each layer is a tool the encoder can switch on; each new family member is the original plus one or two more tools. That layered design is the single most important thing to understand, so hold onto it: every AAC member contains AAC-LC at its core, then adds machinery to squeeze more quality out of fewer bits.

Here is the whole family in one line each, from oldest and simplest to newest and cleverest:

• AAC-LC (Low Complexity) — the plain, full-quality codec. The default for video everywhere.
• HE-AAC v1 (High Efficiency) — AAC-LC plus a treble-rebuilding trick called SBR. Good down to ~48 kbps stereo.
• HE-AAC v2 — HE-AAC v1 plus a stereo-rebuilding trick called Parametric Stereo. Good down to ~24 kbps stereo.
• xHE-AAC (Extended HE-AAC) — a redesigned core that also codes speech, plus mandatory loudness metadata. Good from ~12 kbps mono up to transparent stereo.

The rest of this article unpacks each one, shows the arithmetic that makes the tricks work, and gives you a decision rule. We will lean on the master comparison in how audio compression works for the underlying ideas — psychoacoustic masking, the frequency transform, and entropy coding — so if any of those terms feel shaky, that article is the prerequisite.

Figure 1. One root, three branches. Each AAC member is AAC-LC plus optional tools. The bitrate sweet spot drops as you add tools, because each tool rebuilds something the encoder no longer has to send in full.

AAC-LC: the workhorse you already ship

AAC-LC is the member you are almost certainly using right now without thinking about it. When a phone records a video, when YouTube serves a clip, when an iPhone plays a song from the Music app, the audio is very often AAC-LC. The "LC" stands for Low Complexity, which is a slightly misleading name — it does not mean low quality. It means the codec deliberately leaves out a few computationally expensive tools from the original 1997 AAC design so that cheap hardware can decode it in real time. That trade paid off: AAC-LC decoders are in every phone, browser, smart TV, and set-top box made in the last two decades.

Is AAC the same as AAC-LC? In everyday use, yes: when a spec sheet, phone or encoder says just “AAC”, it almost always means AAC-LC, because AAC-LC is the profile every device decodes and the one most encoders pick by default. Strictly, “AAC” is the whole family and AAC-LC is its base member, which is why “AAC” alone is not a precise enough instruction for an encoder.

The reason AAC-LC replaced MP3 as the default is simple arithmetic dressed up as engineering. At the same bitrate, AAC-LC sounds cleaner than MP3, because it uses a finer frequency analysis and smarter ways to allocate bits to the parts of the sound your ear actually notices. The rule of thumb the industry settled on: AAC-LC reaches "transparent" quality — meaning most listeners cannot tell it from the original — at around 128 kbps for stereo music. MP3 needs roughly 192 kbps to get to the same place. So for the same audio quality, AAC-LC saves about a third of the bytes:

MP3 transparent ≈ 192 kbps
AAC-LC transparent ≈ 128 kbps
192 ÷ 128 = 1.5× → AAC-LC needs about 33% fewer bits

That is why a new product in 2026 that wants maximum compatibility — playable on literally everything — picks AAC-LC for its standard stereo audio. It is the safe default. The standard that defines it is ISO/IEC 14496-3 (MPEG-4 Audio), whose 2019 edition is still the current one as of October 2026; its first amendment, on media authenticity and an immersive interchange format, reached the “under publication” stage at ISO on 29 September 2026 and does not change how AAC-LC is coded. The line traces back to the original AAC in the MPEG-2 standard ISO/IEC 13818-7 from 1997.

Where AAC-LC stops being the right answer is the low-bitrate basement. Below about 96 kbps for stereo, AAC-LC starts to sound thin and grainy, because there simply are not enough bits left to describe the full sound. That is the gap the next two family members were built to fill.

HE-AAC v1: rebuilding the treble you never sent

Imagine you have to describe a photograph to someone over a phone line that is too slow to send the whole picture. You could send the bottom two-thirds in detail and then say, "the top third is more sky, same blue, getting lighter toward the top." The listener paints the missing third from your hint. They do not get the exact original sky, but they get something convincing, and you sent almost no data for it.

That is exactly what HE-AAC v1 does with sound. The trick is called Spectral Band Replication, or SBR. High frequencies — the treble, the sparkle in cymbals and consonants — are expensive to encode in full but are often closely related to the lower frequencies underneath them. So HE-AAC v1 encodes the lower frequencies properly with ordinary AAC-LC, then sends a tiny set of instructions — a few kilobits — telling the decoder how to rebuild the missing treble from the bass and mids it already has. The decoder replicates the upper frequency band, hence the name.

The payoff is large. A treble band that might cost 30 kbps to encode in full can be hinted at for 2–3 kbps. That freed-up budget goes into the part of the sound the ear cares about most. The result: HE-AAC v1 delivers acceptable stereo music at bitrates where plain AAC-LC falls apart — roughly the 48–64 kbps range. The 3GPP standards body's own controlled tests rated HE-AAC and HE-AAC v2 as "Good" for music at bitrates as low as 24 kbps.

HE-AAC v1 began life outside MPEG. A company called Coding Technologies built SBR on top of AAC-LC for satellite radio, marketed it as "aacPlus" and "AAC+", and then handed the SBR method to MPEG, which standardized it in 2003 as ISO/IEC 14496-3:2001/Amd 1:2003. If you ever see a stream labelled "aacPlus", "AAC+", or "eAAC", that is HE-AAC v1 under a vendor trade name. A media-info tool that reports “AAC LC SBR” or “AAC LC + SBR” is describing the same thing: an AAC-LC core with SBR on top, which is HE-AAC v1.

Pitfall — an old decoder hears HE-AAC at half the brightness. Because HE-AAC v1 is AAC-LC plus an SBR hint, a decoder that does not understand SBR will still play the file — it just ignores the treble-rebuild instructions and outputs only the low-frequency AAC-LC core. The audio plays, but it sounds muffled, as if a blanket were thrown over the speakers, and some old players even report the track as twice its real length. If a customer says "your low-bitrate stream sounds dull on my old device", a missing SBR decoder is the prime suspect. Always confirm the target devices support the exact AAC member you ship.

HE-AAC v2: rebuilding the stereo, too

HE-AAC v2 takes the same idea one step further and applies it to the stereo image. Two-channel stereo normally costs nearly twice as much as mono, because you send a left channel and a right channel. But the two channels are usually very similar — most of the sound is in the middle, and the differences between left and right are small and can be described with a handful of numbers rather than a full second channel.

That is Parametric Stereo, or PS. The encoder collapses the two channels down to one mono channel, encodes that mono channel properly, and then sends a tiny side-stream of parameters — numbers that describe how to spread the mono signal back out into a convincing left and right at the decoder. The whole stereo image is rebuilt from one real channel plus a few kilobits of spatial hints.

Stack that on top of SBR and you get HE-AAC v2: AAC-LC for the core, SBR for the treble, Parametric Stereo for the width. Three layers, each rebuilding something the encoder no longer has to send in full. The result is stereo music that stays listenable at around 24–32 kbps — a bitrate at which plain AAC-LC would be unusable noise. Here is the layering as plain arithmetic:

Full stereo, properly coded:        ~128 kbps for transparency
+ SBR (rebuild treble):             treble hinted for ~2–3 kbps instead of ~30 kbps
+ Parametric Stereo (rebuild width): 2nd channel becomes ~3 kbps of parameters
Net result (HE-AAC v2):             listenable stereo at ~24 kbps

HE-AAC v2 was standardized in 2006 as ISO/IEC 14496-3:2005/Amd 2:2006, with the Parametric Stereo tool itself published earlier in 2004. Vendors marketed it as "aacPlus v2", "eAAC+", and "Enhanced AAC+". It became a required audio format in the 3GPP specifications for 3G mobile multimedia, which is why it spread across early mobile streaming and digital radio standards like DAB+ and Digital Radio Mondiale.

There is a quality ceiling, though, and it matters. SBR and Parametric Stereo are reconstructions, not the real thing. They are brilliant when bits are scarce, but as you raise the bitrate they stop helping and can even hold you back, because the reconstructed treble and stereo never quite match a fully coded original. So HE-AAC v2 is a low-bitrate specialist. Above roughly 64 kbps, plain AAC-LC sounds as good or better and is simpler. The encoder, ideally, switches member based on the target bitrate — and that idea of one stream flexing across bitrates is exactly what the newest member was built to deliver.

xHE-AAC: one codec for speech and music, at any bitrate

For forty years, audio engineering kept speech and music in separate boxes. Speech codecs — the ones in your phone calls — model the human voice box and sound superb on voice at tiny bitrates but mangle music. Music codecs like AAC do the opposite: great on music, wasteful on speech. A single product that carries both — say a video with dialogue, then a music sting, then dialogue again — had to either compromise or switch codecs mid-stream.

xHE-AAC ends that split. The "x" stands for Extended, and the codec is built on a technology called USAC — Unified Speech and Audio Coding — standardized as ISO/IEC 23003-3 in 2012. USAC is not AAC-LC with more tools bolted on; it is a redesigned core that can decide, frame by frame, whether a moment of sound is more speech-like or more music-like, and code it with the right tool. The result is a codec that holds up across an enormous bitrate range: roughly 12 kbps for mono speech, climbing smoothly to transparent stereo music at higher rates, all in one continuous family of settings.

Two features make xHE-AAC matter for video products specifically.

The first is bitrate flexibility for adaptive streaming. Modern streaming serves different quality levels to different network conditions — the technique covered in audio adaptive bitrate ladders. xHE-AAC is designed to span that whole ladder with one codec, switching bitrates seamlessly as the network changes, so a phone on a weak connection drops to 16 kbps without an audible break and climbs back to full quality when the signal returns. Netflix reported that when it moved Android mobile streaming to xHE-AAC's adaptive audio, viewers switched from speakers to headphones 16% less often on high-dynamic-range content — a measurable comfort win.

The second is mandatory loudness and dynamic-range metadata, built on a standard called MPEG-D DRC. Every xHE-AAC stream must carry an objective loudness measurement, computed with the ITU-R BS.1770 algorithm explained in loudness, peak, RMS, LUFS, plus instructions the decoder can use to compress the dynamic range on the fly. The practical effect: the listener on a noisy bus can ask their device to flatten the quiet-to-loud range so dialogue stays audible, and the device does it at playback using metadata the encoder already embedded. Loudness control is not an afterthought in xHE-AAC; it is required by the standard.

xHE-AAC is supported in Android since version 9 (Pie, 2018), in iOS since version 13 (2019), and in Windows 11 and Xbox since October 2022. As of Amazon's September 2025 hardware refresh, new Echo and Fire TV devices decode it natively. Fraunhofer lists native decoding across Android (phones, TV, automotive) and Wear OS, every current Apple platform, Amazon Fire and Vega OS, and Windows 11, and names Netflix, Audible, Facebook Stories and Instagram Reels among the services that stream it (2026). The compatibility gap that once made xHE-AAC a risky choice has largely closed for mobile and modern smart-TV audiences.

2026 moved xHE-AAC from “playable” to “producible” on phones. Fraunhofer announced in March 2026 that Android 17 makes xHE-AAC encoding a standard platform feature, not only decoding, across an installed base it puts at more than 7 billion Android devices; the first target is audio messaging under the GSMA RCS Universal Profile, which added xHE-AAC in version 3.1 in 2025. In June 2026 Fraunhofer also reported that Android 17 ships its AAC decoder ported to Rust, with 40–60% lower overall playback complexity than Android 16. On the server side, FFmpeg gained a native AAC USAC (xHE-AAC) decoder in version 7.1 (September 2024), so open-source pipelines can now read xHE-AAC files; encoding still needs a licensed encoder such as Fraunhofer’s.

Picking the right member: a decision you can make in one minute

The choice between AAC members comes down to two questions: how many bits can you spend, and which devices must play it back. The table below is the short version.

Table 1. The four AAC members at a glance. "Sweet spot" is the bitrate range where each member sounds best for its cost; outside that range, a different member usually wins.

A worked example shows how the layers pay off. Suppose you stream a 90-minute talk-and-music show to phones, and you want a fallback rendition for users on a poor cellular connection at 24 kbps stereo. AAC-LC at 24 kbps would be unusable. HE-AAC v2 at 24 kbps is listenable. xHE-AAC at 24 kbps is comfortably good and carries loudness metadata for the noisy-environment listener. If every target phone was sold after 2019, xHE-AAC is the answer. If you must support a fleet of older or cheaper devices, HE-AAC v2 is the safe low-bitrate floor, with AAC-LC for the higher rungs.

Pitfall — "we use AAC" tells your engineers almost nothing. The four members differ by an order of magnitude in their low-bitrate behaviour. A ticket that says "encode the audio in AAC at 32 kbps" without naming the member will get you AAC-LC by default on many tools — and AAC-LC at 32 kbps stereo sounds broken. Always specify the member (AAC-LC, HE-AAC v2, xHE-AAC) and the profile, not just "AAC". The same care applies when you receive files: read the actual profile in the container, because "it's an .m4a" does not tell you which AAC is inside.

How AAC sits inside your files and streams

AAC is a codec, not a file. The encoded AAC bitstream rides inside a container — usually MP4 or its fragmented variant fMP4 for streaming, sometimes a raw ADTS stream with the .aac extension, sometimes Matroska. The container records which AAC profile the bitstream uses, so a correct player reads that field and loads the matching decoder. This is the same codec-versus-container split covered in audio in containers: the container is the box, the AAC profile is what is inside it.

For streaming over HLS and DASH, the profile is also declared in the manifest and in the codecs parameter — strings like mp4a.40.2 for AAC-LC, mp4a.40.5 for HE-AAC v1, mp4a.40.29 for HE-AAC v2, and mp4a.40.42 for xHE-AAC. Players use those strings to decide, before downloading a single audio segment, whether they can play the track. Get the codec string wrong and a capable device may refuse a track it could actually decode, or try to load a track it cannot. How a player picks among audio tracks is covered in depth in audio in HLS, DASH, CMAF.

Reading AAC labels: “AAC 2.0”, “AAC 5.1”, “AAC LC SBR” and levels

“AAC 2.0” means AAC audio with two channels, that is, ordinary stereo. The number after the codec name is the channel layout, not a codec version: “AAC 1.0” is mono, “AAC 2.0” is left and right, and “AAC 5.1” is six channels (front left, centre, front right, two surrounds and a low-frequency channel). There is no “AAC version 2” codec; the closest thing is HE-AAC v2, which is a different member with its own label.

The same labels show up in media-player overlays and in the names of video files. A tag like “MULTI.AAC” or “AAC2.0” in a file name tells you only that the audio is AAC, how many channels it has, and (for “MULTI”) that the file carries several language tracks. It does not tell you the member. To see that, open the file in a media-info tool: “AAC LC” is AAC-LC, “AAC LC SBR” is HE-AAC v1, “AAC LC SBR PS” is HE-AAC v2, and “USAC” is xHE-AAC.

“Level” on a device spec sheet (for example, “HE-AAC Level 2”) is about decoder capacity, not loudness. MPEG-4 Audio profiles come in levels that cap the sampling rate and the number of channels a decoder must handle; a stereo stream needs a lower level than a 5.1 stream. If your product ships surround AAC, check that the target devices list a level that covers your channel count. Audio loudness is a separate setting, handled by the LUFS targets and, in xHE-AAC, by the mandatory DRC metadata.

Licensing in 2026: who pays, and for what

Audio codec licensing confuses people because two separate questions get tangled: do I pay to use the codec software, and do I pay to distribute content encoded with it. For the AAC family the answers, as of 2026, are clearer than the rumours suggest.

The AAC family — AAC-LC, HE-AAC v1, HE-AAC v2, and xHE-AAC, together with the MPEG-D DRC loudness tools — is covered by a single patent pool. That pool was run for years by Via Licensing and, after Via Licensing acquired MPEG LA in 2023, is now administered by Via Licensing Alliance (Via LA). A product company that ships an AAC encoder or decoder — in hardware or in software — needs a license and pays a per-unit royalty. Crucially, xHE-AAC and its DRC tools are included in that same AAC program at no additional cost over baseline AAC, so adopting the newest member does not add a new licensing line item.

The AAC royalty is a per-unit fee on each end-user encoder or decoder product, with the rate falling as annual volume grows. Via LA’s published standard rates as of October 2026:

Table 2. Via LA AAC standard royalty schedule, checked October 2026. A multichannel product counts as 1.5 units; there is also a one-time initial fee of $15,000 ($1,000 for small entities). An alternative rate structure exists; check Via LA for current terms before you budget.

The part that trips people up: unlike the old MP3 program before its 2017 expiry, content owners are not required to pay license fees merely to distribute AAC-encoded content. If you stream AAC audio to your users, you are distributing content, not shipping a codec, and the pool does not charge you for that. The royalty attaches to the encoder and decoder products, which is why the cost usually lives with the operating system, browser, or device maker rather than with you. For most product teams using AAC through the platform's built-in encoder and decoder, the licensing is already handled upstream. If you embed your own commercial AAC encoder, confirm its license terms — but you will not owe a content-distribution fee for the format itself.

This contrasts sharply with Opus, the royalty-free open codec that dominates WebRTC, where there is no pool to license at all. The AAC-versus-Opus choice often comes down to exactly this: AAC's universal device support and broadcast pedigree against Opus's zero licensing and real-time strengths.

Frequently asked questions

What does AAC stand for in audio?

AAC stands for Advanced Audio Coding, the lossy audio codec standardized by MPEG in 1997 as the successor to MP3. It is the default audio format in MP4 files, on Apple devices and in most video streaming.

What is AAC-LC?

AAC-LC (Advanced Audio Coding, Low Complexity) is the base member of the AAC family and the most widely supported audio codec in video. It reaches near-transparent stereo quality at about 128 kbps and is what most devices mean by plain “AAC”.

What is the difference between AAC and AAC-LC?

“AAC” is the name of the whole codec family; AAC-LC is its base profile. In practice, a device or encoder that says only “AAC” is almost always using AAC-LC.

Is HE-AAC better than AAC?

HE-AAC is better than AAC-LC only at low bitrates, roughly below 64 kbps stereo, because SBR and Parametric Stereo rebuild treble and width from a few kilobits. At 96 kbps and above, AAC-LC sounds as good or better and is more compatible.

What does AAC 2.0 mean?

AAC 2.0 means AAC audio with two channels, i.e. stereo. The “2.0” is the channel layout, like “5.1” for surround, not a codec version.

What is xHE-AAC used for?

xHE-AAC is used where one codec has to cover very different bitrates and content: adaptive streaming from about 12 kbps up, mixed speech and music, and services that need built-in loudness control. Netflix, Audible, Facebook and Instagram use it, and Android 17 (2026) makes xHE-AAC encoding a standard platform feature.

Do I have to pay royalties to stream AAC audio?

No. Via LA, which runs the AAC patent pool, charges no fee for distributing AAC-encoded content by streaming, broadcast or physical media. Royalties apply only to encoder and decoder products, from $0.98 per unit down to $0.10 at high volume as of October 2026.

Where Fora Soft fits in

We have built audio into video products across streaming, OTT and Internet TV, video conferencing, telemedicine, e-learning, and AR/VR since 2005. AAC is the audio codec we reach for most in streaming and on-demand work, because it plays on every device a client's audience owns. We pick AAC-LC for the standard rungs of a streaming ladder, layer in HE-AAC or xHE-AAC for the low-bitrate fallback that keeps mobile viewers connected, and lean on xHE-AAC's loudness metadata when a client's content swings between quiet dialogue and loud action. The recurring lesson from production is the one this article opened with: the failure is almost never "AAC is the wrong codec" — it is shipping the wrong AAC member for the bitrate, or never naming the member at all. Our streaming software development team builds AAC ladders with the right member per rung (and, for subscription video, our OTT and IPTV platform developers wire them into the player, DRM and device matrix), and our audio processing engineers handle the transcode, loudness metadata and device testing.

The book

Audio for Video: A Complete Guide to Sound in Video Products: From PCM and Codecs to Loudness, Accessibility, and Real-Time Voice Processing

This article names the four AAC members and tells you which one to ask for; the volume spends 1,229 pages on the ground that choice stands on — sampling, bit depth, channels, loudness and LUFS — and then works through AAC-LC to xHE-AAC, Opus and the speech codecs, with the encoder settings and listening-test evidence behind each bitrate number. Open it when you have to defend a profile, a bitrate and a codec string to the people paying for the bandwidth. Written by Nikolay Sapunov, CEO at Fora Soft.

Read it on Amazon →Kindle & paperback · Fora Soft Video Engineering Handbook

See our case studies

References

1. ISO/IEC 14496-3:2019, Information technology — Coding of audio-visual objects — Part 3: Audio — the current edition of MPEG-4 Audio, the controlling standard for AAC-LC, HE-AAC v1, and HE-AAC v2 profiles. https://www.iso.org/standard/76383.html
2. ISO/IEC 23003-3:2012, Information technology — MPEG audio technologies — Part 3: Unified speech and audio coding (USAC) — the controlling standard for xHE-AAC. https://www.iso.org/standard/57464.html
3. ISO/IEC 14496-3:2001/Amd 1:2003 — the amendment that first standardized the HE-AAC v1 (SBR) profile. https://www.iso.org/standard/38148.html
4. ISO/IEC 14496-3:2005/Amd 2:2006 — the amendment that first standardized the HE-AAC v2 (SBR + Parametric Stereo) profile. https://www.iso.org/standard/43026.html
5. Fraunhofer IIS, xHE-AAC product and technology pages — the codec's maker on USAC, MPEG-D DRC loudness, and platform support. https://www.iis.fraunhofer.de/en/ff/amm/broadcast-streaming/xheaac.html
6. Fraunhofer Audio Blog, Netflix Now Streaming xHE-AAC Audio on Android Mobile (2021) — primary report of Netflix's adaptive-audio adoption and the 16% headphone-switching figure. https://www.audioblog.iis.fraunhofer.com/netflix-xheaac-android
7. Fraunhofer Audio Blog, xHE-AAC Audio Codec supported in Amazon's new Line of Products (2025) — Amazon Echo / Fire TV native xHE-AAC support. https://www.audioblog.iis.fraunhofer.com/xhe-aac-amazon-vega
8. High-Efficiency Advanced Audio Coding — overview of HE-AAC v1/v2 standardization history, trade names, and the Via Licensing patent pool; used for orientation, with every standards claim confirmed against the ISO/IEC amendment listings in references 3–4. https://en.wikipedia.org/wiki/High-Efficiency_Advanced_Audio_Coding
9. EBU Technical Review, Moser, MPEG-4 HE-AAC v2 — audio coding for today's media world (2006) — the EBU's explanation of SBR and Parametric Stereo. https://tech.ebu.ch/docs/techreview/trev_305-moser.pdf
10. Via LA Licensing, AAC patent pool program — administrator of the AAC family license formed when Via Licensing acquired MPEG LA in 2023. https://www.via-la.com/
11. ISO/IEC 14496-3:2019/Amd 1, Media authenticity and immersive interchange format — stage 60.00 (under publication) on 29 September 2026. https://www.iso.org/standard/92052.html
12. Fraunhofer Audio Blog, xHE-AAC Audio Encoding becomes a Standard Feature in Android 17 (March 2026). https://www.audioblog.iis.fraunhofer.com/xhe-aac-android-17
13. Fraunhofer Audio Blog, World’s First Rust Codec Implementation for Google Android (June 2026) — the Android 17 AAC decoder and its 40–60% complexity reduction. https://www.audioblog.iis.fraunhofer.com/worlds-first-rust-codec-implementation-for-google-android
14. Fraunhofer Audio Blog, xHE-AAC: A Sound Upgrade for SMS Messages (August 2025) — xHE-AAC in GSMA RCS Universal Profile 3.1. https://www.audioblog.iis.fraunhofer.com/xhe-aac-rcs_3-1
15. FFmpeg project news, FFmpeg 7.1 “Péter” (30 September 2024) — native AAC USAC (xHE-AAC) decoder. https://ffmpeg.org/index.html
16. Via LA, AAC licensing program — covered technologies, no fee on content distribution, standard royalty schedule (checked October 2026). https://www.via-la.com/licensing-programs/aac/

Per §4.3.2, where the popular "AAC is licensed for content distribution" framing (common on SEO blogs, reference 8's secondary citations) conflicts with the patent pool's own terms, this article follows the pool's published position (reference 10) — content distribution carries no fee; the royalty attaches to encoders and decoders.