Skip to content

CONCEPT Cited by 1 source

Split-band audio coding

Definition

Split-band (aka subband) coding splits the input audio spectrum into two or more frequency bands, encodes each band separately, and recombines at the decoder. Bands can use different bitrate budgets and different encoders, and can share parameters / excitation information across bands for better overall compression (Source: sources/2024-06-13-meta-mlow-metas-low-bitrate-audio-codec).

In MLow

Per the post's block diagram description:

"On the left we have an input signal (raw PCM audio) feeding into the encoder, which then splits the signal into two low and high-frequency bands. Then, each band is encoded separately while making use of shared information to achieve better compression."

Two bands + cross-band information sharing + range encoder on the merged output. The decoder inverts.

Why it helps at low bitrate

Low and high frequency bands have different information content:

  • Low band (voice fundamentals + low harmonics) carries most perceptually-important energy — worth spending bitrate on.
  • High band (sibilants, air, room texture) can often be reconstructed from low-band parameters with a very small bitrate supplement — spectral envelope + excitation hints rather than full waveform.

The payoff Meta claims:

"With these split-band optimizations, we are able to encode the high band using very few bits, which lets MLow deliver SuperWideBand (32kHz sampling) using a much lower bitrate."

I.e., the band / bitrate coupling that caps Opus at NarrowBand at 6 kbps is relaxed by smart band partitioning — see concepts/narrowband-vs-wideband-audio.

Broader technique family

Split-band coding is the operating principle of older codecs like G.722 (ADPCM sub-band) and the SILK half of Opus. The MLow contribution is not introducing the family but refining the information-sharing + quantization + range-coding tail on top of a CELP foundation.

Seen in

Last updated · 319 distilled / 1,201 read