Should the Auracast stream be delayed to line up with the loudspeakers?

Why it matters

Sound from loudspeakers travels at about 343 m/s, so it reaches a listener 30 m away roughly 90 ms after leaving the loudspeaker¹. Auracast adds device latency of its own, so the stream can arrive before or after the room sound^2,3. Listeners who hear both, through open-fit hearing aids, a program that mixes the stream with the hearing aid's microphones, or the transparency mode of their earbuds, hear two copies of the same programme at different times^1,4. Whether the assistive signal should be delayed to match the room, and by how much, changes how a system is designed and set up¹.

What the evidence shows

• The usual tolerance figure comes from hearing-aid delay research. A 2023 position paper classifies a second copy below about 5 ms as coloration, 5 to 30 ms as reverberation and above 30 ms as an echo, and recommends delay compensation wherever any seat is more than about 10 m from the stage¹. The tolerance study it cites measured a hearing-device processing delay, where the processed copy arrives after the unprocessed sound, using a laboratory setup that the authors say differed from commercial hearing aids⁵.
• In that situation the late, louder copy dominates. In a DFG-funded study with 13 normal-hearing participants (11 completed both sessions), with a louder copy delayed by up to 7 ms in one ear, the median speech reception threshold in noise with simulated hearing loss worsened by about 3.4 dB as the delay grew (from -18.1 to -14.7 dB SNR); the authors attribute this to backward masking. The same delay in both ears produced no significant change⁶.
• People with hearing loss tolerate more delay. In a Southampton study funded by EPSRC and EU grants, 20 listeners with mild to moderate hearing loss tolerated more processing delay than 20 normal-hearing listeners in two of three voice conditions, and the greater their loss, the less their annoyance rose with delay⁵; a separate independent university study of 9 listeners with hearing loss and 8 with normal hearing also found higher echo thresholds with hearing loss for short noise bursts⁷.
• Open fittings already mix two copies. With open-fit hearing aids, room sound passing through the vent mixes with the delayed device output and causes comb filtering⁴. As evidence about this shared mixing path, a bench study of two proprietary remote-microphone systems, with total delays of about 23-24 ms and 33-35 ms, found comb-filter ripples when the signals were mixed at similar gain with the hearing aid's own microphones; it tested no listeners or venue systems. The authors report no external research funding; IRCCS Burlo Garofolo paid the publication charge, and two authors worked for a hearing-care company⁸.
• Early reflections after direct sound can help speech. In independent research by the National Research Council Canada and Université de Montréal with 21 listeners who did not use hearing aids, early reflections arriving within 50 ms of the direct sound increased effective signal-to-noise ratio and intelligibility as much as the same added direct sound, both for listeners with near-normal hearing and for older listeners with mild to moderate hearing loss⁹.
• Lip-reading tolerates late audio much better than early audio. In an NSF-funded test with nine normal-hearing listeners and filtered speech, audio leading the talker's face by 40 ms, the shortest lead tested, reduced intelligibility, while with audio lagging, intelligibility stayed relatively stable up to about 160 to 200 ms¹⁰. In a study from the manufacturer Starkey's research centre, the average voice-face synchrony window did not differ between 11 listeners with moderate to severe hearing loss and 11 normal-hearing listeners; individual windows varied widely and were on the order of a few hundred milliseconds¹¹. For television, general audiences detected sound leading the picture by about 45 ms and lagging it by about 125 ms¹².
• Auracast is not instantaneous. Pro Audito's independent tests, with devices lent free by manufacturers, measured transmitter-to-listener latency from 26 to 205 ms depending on the pair². Field and product tests measured 36 to 43 ms to receivers with neckloops and 56 to 58 ms to hearing aids in a theatre¹³, 43 ms to a receiver and 54 ms to a hearing aid outdoors¹⁴, and 65 to 75 ms through hearing aids from a mid-priced transmitter¹⁵. Near the stage the stream can therefore arrive after the loudspeaker sound; far from the stage the two can roughly coincide.
• Makers' and advocates' figures differ. Transmitter makers state 29 to 45 ms, or under 40 ms, with their own receivers only^16,17. The Center for Hearing Access gives about 36 to over 100 ms for Auracast, against about 1 to 30 ms for loops, FM and infrared¹⁸.
• The specification requires support for 40 ms. Every broadcast setting in the table lists a 40 ms Presentation Delay as a receiver capability requirement. Total system delay also includes radio transport and the transmitter's own processing time; this is not a fixed end-to-end latency¹⁹. The SIG's September 2026 transmitter guide recommends the 24 kHz low-latency setting for live use and allows professionally installed transmitters to let installers adjust Presentation Delay for larger venues²⁰.
• The whole chain counts. In the theatre trial, the audio system and the sound path to a distant microphone added about 30 ms, for a total of roughly 65 to 90 ms¹³. The Center for Hearing Access treats latency as the sum of every link from microphone to hearing instrument²¹, and total delay through hearing aids and headphones is often about 70 ms²². US rules require prescription hearing aids to state a latency value in their labelling, measured to within 1.5 ms, and limit over-the-counter hearing aids to 15 ms; neither rule deals with latency over a wireless link²³.
• Far from the stage, the stream can lead. At a rear seat in a Swiss arena, the Auracast signal arrived less than 30 ms ahead of the loudspeaker sound, with smaller offsets nearer the loudspeakers³. In an AGE-WELL-funded concert-hall study with 20 study-aid users and 8 own-aid users, with hearing aids and fittings supplied by Sonova Canada, the loop feed was delayed to match the direct sound at the looped seats, and the authors suggest that about 40 ms of uncorrected delay may help explain the lower ratings of another feed²⁴.
• Listeners are spread out. A loop installer can set one delay for the looped area, whereas Auracast listeners are not confined to one position and so experience a range of acoustic delays²⁵.

Where sources disagree

One view holds that alignment compensation is a requirement for a good experience beyond about 10 m¹. Another reading draws on room-acoustics and echo research: early reflections within 50 ms of the direct sound aided intelligibility⁹, and listeners with hearing loss heard a short delayed copy as a separate echo only at longer delays than normal-hearing listeners⁷. Neither study tested a stream that arrives before the room sound.

Guidance documents treat 40 ms as the maximum acceptable difference from the room sound^22,26,27, two of them with 30 ms preferred^26,27, but do not say which signal should come first. Both the UK code and the Swiss draft limit their 30/40 ms guidance to audible speech signals of similar level; the UK code notes that percussive music may need a shorter delay^26,27. The Center for Hearing Access applies a similar 30 to 40 ms rule of thumb, marked as under review, to end-to-end latency instead²¹, and a book distributed by the SIG states, without citing a study, that a gap of much more than 30 to 40 ms begins to be heard as echo²⁸.

Tests also disagree on which devices are slower: hearing aids tended to be slower than other receivers in one test², while earbuds were generally much slower than hearing aids in a McMaster/SoundsGood Labs synopsis that gives neither the device count nor a funding statement²⁹.

Field testers reached different conclusions at different positions: in a theatre, 56 to 58 ms to hearing aids was judged too long, and a total of 65 to 90 ms problematic for some seats near the front¹³; at an arena's rear seat, where the stream arrived ahead of the loudspeaker sound, the authors concluded that no disturbing echo arises³.

What would settle it

A blind listening test with hearing-aid and implant users, in a real reverberant room, comparing an assistive signal that leads the room sound by several amounts with a time-aligned one. It should measure speech understanding and listening effort as well as preference, and include open and closed fittings with microphones on and off. These would address the different signal orders, listener groups and test conditions in the evidence reviewed here^5,9,24.

Planning while it is open

• Expect Auracast itself to add latency, so check the timing near the stage, where the stream is most likely to arrive after the loudspeakers^22,13.
• Measure latency through hearing aids and earbuds as well as dedicated receivers; latency differed by device type, in opposite directions in different tests^2,13,29.
• For live use, the SIG recommends the 24 kHz low-latency setting²⁰.
• Where seats are far apart, separate broadcasts with different delays are one proposed option¹; receivers are not required to support Presentation Delays above 40 ms, so many may not be able to buffer enough to line up with the room sound in large venues¹. The SIG allows professionally installed transmitters to offer an adjustable Presentation Delay²⁰.
• Where video screens carry the talker, compare the stream with the picture as well as with the room: the UK code asks for audio-to-picture timing preferably under 40 ms and no more than 50 ms²⁶, and a filtered-speech test found reduced intelligibility at an audio lead of 40 ms, its shortest tested lead¹⁰.

Sources

1. Requirements for mass adoption of assistive listening technology by the general public. Kaufmann TB, Foroogozar M, Liss J, Berisha V. ICASSP 2023 Workshops, IEEE (arXiv 2303.02523), 3 May 2023. Independent (Arizona State University) ↩︎ https://arxiv.org/abs/2303.02523
2. Auracast transmitter-receiver compatibility test (with data workbook). Pro Audito Schweiz, 6 October 2025. Independent (Pro Audito Schweiz, Swiss hearing-loss association; no manufacturer funding, test products lent free by manufacturers) ↩︎ https://pro-audito.ch/news/auracast-transmitter-receiver-compatibility-test/
3. Technical analysis of the Bluetooth Auracast system in the Pilatus Arena. Pro Audito Schweiz. Pro Audito Schweiz (in German), 3 November 2025. Independent (Pro Audito Schweiz, Swiss hearing-loss association; test devices were lent free by their manufacturers or owned by participants) ↩︎ https://pro-audito.ch/news/technische-analyse-des-bluetooth-auracast-systems-in-der-pilatus-arena/
4. Hearing aids: what works well and what can be improved. Moore BCJ. Journal of the Association for Research in Otolaryngology 27:123-135, 13 February 2026. Independent (University of Cambridge; UK research council grant) ↩︎ https://link.springer.com/article/10.1007/s10162-026-01031-5
5. Tolerable delay for speech production and perception: effects of hearing ability and experience with hearing aids. Goehring T, Chapman JL, Bleeck S, Monaghan JJM. International Journal of Audiology 57(1):61-68, 2018. Independent (University of Southampton; EPSRC and EU Seventh Framework Programme grants) ↩︎ https://pubmed.ncbi.nlm.nih.gov/28838277/
6. Effect of a processing delay between direct and delayed sound in simulated open fit hearing aids on speech intelligibility in noise. Roth S, Müller F-U, Angermeier J, Hemmert W, Zirn S. Frontiers in Neuroscience 17:1257720, 4 January 2024. Independent (Offenburg University of Applied Sciences and Technical University of Munich; German Research Foundation, DFG) ↩︎ https://pmc.ncbi.nlm.nih.gov/articles/PMC10805375/
7. Effects of hearing loss on echo thresholds. Roberts RA, Besing J, Koehnke J. Ear and Hearing 23(4):349-357, August 2002. Independent (University of South Florida) ↩︎ https://pubmed.ncbi.nlm.nih.gov/12195177
8. Time delay and frequency analysis of remote microphones. Andreatta E, Caregnato I, Selmo A, Gulli A, Onofrei MG, Orzan E. Audiology Research 15(5):123, 25 September 2025. Independent (authors report no external research funding; IRCCS Burlo Garofolo paid the publication charge through Ricerca Corrente RC 05/21); two authors employed by Acustica Caregnato ↩︎ https://www.mdpi.com/2039-4349/15/5/123
9. On the importance of early reflections for speech in rooms. Bradley JS, Sato H, Picard M. Journal of the Acoustical Society of America 113(6):3233-3244, June 2003. Independent (National Research Council Canada; Université de Montréal) ↩︎ https://nrc-publications.canada.ca/eng/view/accepted/?id=dd3c1e7a-8d8d-440e-bbcd-04c69ef20419
10. Speech intelligibility derived from asynchronous processing of auditory-visual information. Grant KW, Greenberg S. Proceedings of Auditory-Visual Speech Processing (AVSP) 2001, pp. 132-137 (ISCA Archive), September 2001. Independent (Walter Reed Army Medical Center and International Computer Science Institute; US National Science Foundation grant) ↩︎ https://www.isca-archive.org/avsp_2001/grant01_avsp.html
11. Audiovisual asynchrony detection and speech intelligibility in noise with moderate to severe sensorineural hearing impairment. Başkent D, Bazo D. Ear and Hearing 32(5):582-592, September 2011. Manufacturer (Starkey Hearing Research Center affiliation) ↩︎ https://pubmed.ncbi.nlm.nih.gov/21389856/
12. Relative timing of sound and vision for broadcasting (Recommendation ITU-R BT.1359-1). International Telecommunication Union, Radiocommunication Sector (ITU-R). ITU-R, November 1998. Standards body (International Telecommunication Union, Radiocommunication Sector) ↩︎ https://www.itu.int/rec/R-REC-BT.1359-1-199811-I/en
13. Test report: Auracast assistive listening system in a theatre. Pro Audito Schweiz. Pro Audito Schweiz (in German), 17 March 2025. Independent (Pro Audito Schweiz, Swiss hearing-loss association) ↩︎ https://pro-audito.ch/news/testbericht-auracast-hoeranlage-im-theater/
14. Auracast practical test outdoors. Pro Audito Schweiz. Pro Audito Schweiz; English reprint by the Center for Hearing Access, 16 September 2026. Independent (Pro Audito Schweiz, Swiss hearing-loss association) ↩︎ https://centerforhearingaccess.org/wp-content/uploads/Pro-Audito-Schweiz-Auracast-practical-test-outdoors.pdf
15. Tested for you: the Avantree Venucast Oasis Aura Auracast transmitter (in German). Pro Audito Schweiz, 16 July 2026. Independent (Pro Audito Schweiz, Swiss hearing-loss association; test units supplied free by the manufacturer, Pro Audito not paid) ↩︎ https://pro-audito.ch/news/fuer-sie-getestet-auracast-sender-avantree-venucast-oasis-aura/
16. Auri TX2N / TX2N-D transmitter datasheet. Listen Technologies and Ampetronic (Auri). Ampetronic, July 2025. Manufacturer (Listen Technologies and Ampetronic) ↩︎ https://www.ampetronic.com/wp-content/uploads/UP3B603-2_Auri_TX2N_Transmitter_Datasheet.pdf
17. Infinium system controller and transmitter user manual (MAN 297B). Williams AV, 2026. Manufacturer (Williams AV) ↩︎ https://williamsav.com/wp-content/uploads/2025/08/MAN-297B-Infinium-Controller-Transmitter-BA-CT1.pdf
18. Assistive Listening Systems Quick Guide. Center for Hearing Access (reviewed by Juliette Sterkens, AuD). Center for Hearing Access, The John G. Shedd Institute, 29 October 2025. Nonprofit advocacy (Center for Hearing Access, an initiative of The John G. Shedd Institute for the Arts; it states that it accepts no donations from hearing-related vendors) ↩︎ https://centerforhearingaccess.org/wp-content/uploads/Assistive-Listening-Systems-Quick-Guide.pdf
19. Basic Audio Profile, version 1.0.3. Bluetooth SIG Generic Audio Working Group. Bluetooth SIG, 23 June 2026. Standards body (Bluetooth SIG, an industry association funded by its member companies, which owns the Auracast trademark) ↩︎ https://www.bluetooth.com/wp-content/uploads/Files/Specification/HTML/bap_v1-0-3_1782503972/BAP_v1.0.3/out/en/index-en.html
20. How to build an Auracast transmitter: technical overview, version 3. Bluetooth SIG Market Development. Bluetooth SIG, 16 September 2026. Standards body (Bluetooth SIG, an industry association funded by its member companies, which owns the Auracast trademark) ↩︎ https://www.bluetooth.com/wp-content/uploads/2024/05/2403_How_To_Auracast_Transmitter.pdf
21. What is needed for Auracast streamed assistive listening systems to be available and usable. Center for Hearing Access (reviewed by Juliette Sterkens, AuD). Center for Hearing Access, The John G. Shedd Institute, 15 December 2024. Nonprofit advocacy (Center for Hearing Access, an initiative of The John G. Shedd Institute for the Arts; it states that it accepts no donations from hearing-related vendors) ↩︎ https://centerforhearingaccess.org/wp-content/uploads/What-is-needed-for-Auracast-streamed-ALS-to-be-available-and-usable.pdf
22. Bluetooth Auracast in theaters and event venues. Graf B (Pro Audito Schweiz). Proscenium no. 3, pp. 10-11, Swiss Association of Technical Stage and Event Industry; English reprint by the Center for Hearing Access, September 2026. Nonprofit advocacy (Pro Audito Schweiz, Swiss hearing-loss association) ↩︎ https://centerforhearingaccess.org/wp-content/uploads/Proscenium-Technical-article-Bluetooth-Auracast-in-theaters-and-event-venues-1.pdf
23. 21 CFR 801.422: Prescription hearing aid labeling. US Food and Drug Administration. Electronic Code of Federal Regulations, Title 21, 17 October 2022. Government (US Food and Drug Administration) ↩︎ https://www.ecfr.gov/current/title-21/chapter-I/subchapter-H/part-801/subpart-H/section-801.422
24. Evaluating a hearing loop implementation for live orchestral music. McWeeny S, Trainor LJ, Armstrong S, Bosnyak D, Tawfik H, Bruce IC. Ear and Hearing 46(3):808-816, 27 January 2025. Independent (McMaster University; AGE-WELL Canada grant); in-kind hearing aids and fittings from Sonova Canada ↩︎ https://pubmed.ncbi.nlm.nih.gov/39868727/
25. Opportunities and challenges for Bluetooth LE Audio assistive listening systems. Bruce IC, Armstrong S, Bosnyak DJ, Tawfik H. IEEE ICASSP 2025 (McMaster University), 12 March 2025. Independent (McMaster University); funding not stated in the abstract ↩︎ https://doi.org/10.1109/ICASSP49660.2025.10889430
26. Code of Practice for the Design and Operation of Assistive Listening Systems, Issue 1. Audio Industries Committee on Assistive Listening Systems. Audio Industries Committee on Assistive Listening Systems; distributed by the Institute of Sound, Communications and Visual Engineers (UK), December 2023. Trade body (Audio Industries Committee on Assistive Listening Systems, an audio industry committee; distributed by the Institute of Sound, Communications and Visual Engineers; the code names no funder) ↩︎ https://iscve.org.uk/standards/assistive-listening-guide-and-code-of-practice/
27. prSIA 500:2025-07 Hindernisfreie Bauten (obstacle-free buildings), consultation draft. SIA (Swiss Society of Engineers and Architects). SIA; consultation copy with highlights published by Pro Audito Schweiz (in German), July 2025. Standards body (SIA) ↩︎ https://pro-audito.ch/wp-content/uploads/prSIA500_Vernehmlassungsentwurf-mit-Hervorhebungen.pdf
28. Introducing Bluetooth LE Audio, 2nd edition (including Auracast). Nick Hunn. Nick Hunn; free digital edition distributed by the Bluetooth SIG, November 2024. Independent (the author's own book; free edition distributed by the Bluetooth SIG; the 2024 Broadcast Audio URI specification lists the author under GN Hearing, a hearing-aid maker) ↩︎ https://www.bluetooth.com/wp-content/uploads/2025/01/Introduction-to-LE-Audio-2.0.pdf
29. Evaluation of the first generation of Auracast-compatible transmitters, hearing aids and earbuds. Bruce I, Clarke S, Armstrong S, Bosnyak D. IHCON 2026 (McMaster University; SoundsGood Labs), August 2026. Independent (McMaster University with SoundsGood Labs, which the synopsis does not describe; no funding statement in the synopsis) ↩︎ https://centerforhearingaccess.org/wp-content/uploads/IHCON-Synopsis-Bruce-Evaluation-of-the-first-generation-of-Auracast-compatible-transmitters-hearing-aids-and-Earbuds-August-2026.pdf

Change log

• 10 October 2026Published with 29 sources.
• 13 October 2026Qualified the evidence gap and timing requirements; added study samples, funding and test limits.
• 13 October 2026Corrected research funding and qualified synchrony and speech-delay results against originals.
• 14 October 2026Prepared a table-linked evidence block for owner review; the block remains unpublished.
• 14 October 2026Audited the unpublished evidence block; corrected the Aarau audience and timing scope and the Auri source locator.
• 14 October 2026Added Evidence at a glance, drawn from the data tables.

Cite this question

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, or see how suggestions are handled.