Assistive Listening Knowledge Base
  • White paper

Performance characterization of the Low Complexity Communication Codec

From https://assistivelisteningknowledgebase.org/evidence/bluetooth-sig-2023-lc3-characterization/ · record checked 13 October 2026

  • Bluetooth SIG Generic Audio Working Group
  • Bluetooth SIG

Standards body (Bluetooth SIG, an industry association funded by its member companies, which owns the Auracast trademark; laboratory tests by SenseLab of FORCE Technology and by Fraunhofer IIS)

Summary

Summarises the listening tests and measurements made while the LC3 codec used by Auracast was standardised.

Key findings

  • In the hearing-aid speech test at 16 kHz, LC3 at 32 kbps was rated the same as SBC at 66.4 kbps and significantly better than G.722 at 64 kbps (online test, 119 listeners).
  • In the hearing-aid music test at 24 kHz, LC3 at 48 kbps was rated significantly better than G.722 at 96 kbps and SBC at 96.4 kbps (online test with 115 listeners, and a laboratory test with 20 listeners).
  • The hearing-aid-use-case listening tests in section 2.4 used error-free transmission. Section 1.3.5 allows implementers to modify or replace the example packet-loss concealment algorithm; section 2.1.2 separately reports a Fraunhofer test with 25 listeners comparing concealment implementations at loss rates from 0% to 12%. The section 2 introduction warns that transmission errors can affect listening quality.
  • Section 3.2 (pages 38-39) says LC3 uses a long-term post-filter at lower bitrates. For level and distortion measurements, this filter can raise a full-scale sine tone by up to 0.5 dB; a 6,427 Hz test tone usually keeps the gain difference below 0.03 dB at 16, 24, 32 and 48 kHz, and for normal speech and music the difference is typically under 0.1 dB.
  • Section 1.3.4 explains that LC3 codes channels separately and builds stereo or multichannel streams by aggregating those channels. Codec payload bitrate therefore scales with channel count.
  • Section 2.4.4 reports a second 16 kHz test with 20 expert listeners: 10 ms LC3 at 32 kbps rated significantly above 7.5 ms LC3 at that rate; 7.5 ms LC3 did not differ significantly from G.722 at 64 kbps.
  • Tables 2.6-2.8 and sections 2.4.3-2.4.6 identify both speech and music in the 16 kHz and 24 kHz material. The online LC3 conditions used 10 ms frames; the laboratory conditions included both 7.5 ms and 10 ms frames.

Limits

Compares codecs, not complete broadcast systems. The paper does not report the listeners' hearing; the online tests were open to Bluetooth SIG members.

Read the original

www.bluetooth.com/wp-content/uploads/2023/07/LC3Characterization_WP.pdf

Link checked . We link to the original and do not host copies.

Cited in

Cite this source

Bluetooth SIG Generic Audio Working Group. Performance characterization of the Low Complexity Communication Codec. Bluetooth SIG. 13 June 2023. https://www.bluetooth.com/wp-content/uploads/2023/07/LC3Characterization_WP.pdf (accessed 13 October 2026).

The accessed date is when this library last checked the link; use the date you read it. To point readers to this summary instead, link to https://assistivelisteningknowledgebase.org/evidence/bluetooth-sig-2023-lc3-characterization/

Every finding here cites its source, and who funded each source is labelled. This site does not recommend products. Found an error or a missing study? Suggest a correction.