HE-AAC v2 Audio
HE-AAC v2 adds parametric stereo coding to HE-AAC for efficient delivery of stereo audio at very low bitrates.
A stereo radio programme is delivered with a mono core, yet a full decoder produces stereo. HE-AAC v2's Parametric Stereo (PS) tool describes the stereo relationship alongside a coded mono downmix. The output channels are reconstructed using both.
A mono combination alone is insufficient
If a downmix contains the average of L and R, the pairs (0.8,0.2) and (0.5,0.5) both give 0.5. Their stereo impressions differ, but the combined value does not identify which pair existed.
PS sends information about the relationships that matter for reconstructing the stereo image. Its parameters describe properties such as relative channel energy and correlation in time/frequency regions. A decoder uses these descriptions to distribute and shape the reconstructed signal into two outputs.
| What survives in the delivery | What the decoder can do |
|---|---|
| Mono downmix alone | Reproduce the combined signal |
| Downmix plus stereo parameters | Reconstruct the described stereo relationships |
| Original separate sample sequences | Preserve more independent channel detail, at a different coding cost |
The PS representation is lossy. Parameters describe aspects of the relationship rather than carrying a second complete waveform that uniquely restores every original left/right sample.
From the AAC core to full-band stereo
For a stereo presentation, the chain combines three jobs:
- The AAC core reconstructs its coded lower-band signal.
- Spectral Band Replication reconstructs the intended wider audio band.
- Parametric Stereo produces the stereo image from that signal and the PS information.
The HE-AAC article explains the first two jobs and their possible core/output-rate difference. PS adds a channel-reconstruction task, not a different meaning for the sample timeline.
A suitably compatible core-only decode can produce narrower-band audio. SBR without PS can restore the wider band but leave the mono presentation. Full stereo reconstruction requires the PS information to be used. A report showing a mono core is therefore consistent with a correctly reconstructed stereo programme.
Spend a small budget on useful relationships
At restricted stereo bitrates, coding one main signal plus parameters can be more efficient than coding two complete HE-AAC signals. That can retain a useful stereo impression in broadcasts.
The compromise becomes important when independent or rapidly changing channel detail is difficult to describe parametrically. A stereo result can be recognizable without matching every original spatial nuance.
The HE-AAC v2 profile also permits multichannel configurations where PS is not used; those use the relevant HE-AAC tools instead. For an Emby recording, read its actual presentation and tool use rather than turning the profile label into a claim that every v2 track was encoded as mono-plus-PS.