# Why is the sound quality of onboard public address so bad?

While not a frequent flyer, I did fly common airliners quite a lot, and each time the (very) low sound quality of onboard public adress system has always struck me.

We all know technology makes it possible to have a "HiFi" sound quality almost everywhere, so why don't major manufacturers (Boeing and Airbus) provide better sound quality systems?

While this is of course not mandatory to make an airplane fly, it is related to passenger safety. Better sound quality would allow crew instructions to be better understood during the pre-takeoff briefing and in the event of an emergency.

I have flown with several airlines, and there are already enough other factors that make it difficult to understand what's being said on PA:

• On board noise: air conditioning, running engines, neighbors talking, kids crying, ...
• Strong or difficult english accent when either the crew or the passenger are not english-native speakers.
• Crew speaking really fast.

The airplane companies can't do much on the above items, but they can improve the technical aspects. Of course, I do understand that this would require increasing weight (loudspeakers are heavy...) thus reducing commercial load.

Side question: How do US/EU regulations handle this point? Do they have any specs on sound quality requirements?

Note: I haven't tried recent planes such as A380, so maybe it has been improved.

• On the point about a bad accent, I've been on a few different European airlines. They usually speak live in their native language (German, Norwegian, French, etc) and then play an English MP3 using the FAP. Or if the crew is English, vice versa. – Danny Beckett May 30 '15 at 12:28
• Crews speaking really fast: I notice this, and find it very annoying. Crews must have made the same announcement at least thousands of times, but passengers need to hear it! – kevin May 30 '15 at 13:48
• United Kingdom civil aviation authority's specification on PAS – anshabhi May 30 '15 at 13:49
• Thanks @anshabhi, so this UK regulation indeeds requires a comprehension rate that needs to be achieved. – kebs May 30 '15 at 16:09
• Probably for the same reason the salads are so bad. – Tyler Durden May 31 '15 at 2:37

For the regulatory portion of this question.

14 CFR 121.318

No person may operate an airplane with a seating capacity of more than 19 passengers unless it is equipped with a public address system which—

(a) Is capable of operation independent of the crewmember interphone system required by§ 121.319, except for handsets, headsets, microphones, selector switches, and signaling devices;

(b) Is approved in accordance with§ 21.305 of this chapter;

(c) Is accessible for immediate use from each of two flight crewmember stations in the pilot compartment;

(d) For each required floor-level passenger emergency exit which has an adjacent flight attendant seat, has a microphone which is readily accessible to the seated flight attendant, except that one microphone may serve more than one exit, provided the proximity of the exits allows unassisted verbal communication between seated flight attendants;

(e) Is capable of operation within 10 seconds by a flight attendant at each of those stations in the passenger compartment from which its use is accessible;

(f) Is audible at all passenger seats, lavatories, and flight attendant seats and work stations; and

(g) For transport category airplanes manufactured on or after November 27, 1990, meets the requirements of§ 25.1423 of this chapter.

[Doc. No. 24995, 54 FR 43926, Oct. 27, 1989]

14 CFR 25.1423

A public address system required by this chapter must—

(a) Be powerable when the aircraft is in flight or stopped on the ground, after the shutdown or failure of all engines and auxiliary power units, or the disconnection or failure of all power sources dependent on their continued operation, for—

(1) A time duration of at least 10 minutes, including an aggregate time duration of at least 5 minutes of announcements made by flight and cabin crewmembers, considering all other loads which may remain powered by the same source when all other power sources are inoperative; and

(2) An additional time duration in its standby state appropriate or required for any other loads that are powered by the same source and that are essential to safety of flight or required during emergency conditions.

(b) Be capable of operation within 3 seconds from the time a microphone is removed from its stowage.

(c) Be intelligible at all passenger seats, lavatories, and flight attendant seats and work stations.

(d) Be designed so that no unused, unstowed microphone will render the system inoperative.

(e) Be capable of functioning independently of any required crewmember interphone system.

(f) Be accessible for immediate use from each of two flight crewmember stations in the pilot compartment.

(g) For each required floor-level passenger emergency exit which has an adjacent flight attendant seat, have a microphone which is readily accessible to the seated flight attendant, except that one microphone may serve more than one exit, provided the proximity of the exits allows unassisted verbal communication between seated flight attendants.

[Doc. No. 26003, 58 FR 45229, Aug. 26, 1993, as amended by Amdt. 25-115, 69 FR 40527, July 2, 2004]

There are systems in place to increase volume in certain instances, eg, on the 767 both engines running gives a 6 dB increase and O2 masks dropped gives an additional 3 dB gain in volume.

• A good percentage of the flights I've been on failed (c), and I would expect that virtually all commercial aircraft would fail that requirement when ambient noise levels are high, which would often be the case in emergencies when a PA system is needed most. I think it just underscores the validity of the OP's question. – Carey Gregory May 31 '15 at 17:32
• Not disputing the awfulness of the typical PA system. From DC-9 to B777 they're all pretty bad. There are systems in place to increase volume in certain instances, eg, on the 767 both engines running gives a 6 dB increase and O2 masks dropped gives an additional 3 dB gain in volume. This probably really just means louder intelligibility. – Sports Racer May 31 '15 at 18:03

The PA (Public Address) system is composed by a band-pass filter between a range of frequencies $f \in [f_i , f_f]$

The quality of the sound is affected because the voice frequencies may exceed this range and thus, the band-pass filter rejects these peaks, that are exceeding the filter range.

Remember how a band-pass filter works:

The frequencies below $f_i$ are not allowed nor the frequencies above $f_f$. Nevertheless, when the frequencies are outside of the bandwidth range (determined by the $\Delta f=f_2-f_1$ at which the gain has a value of $\mathrm{gain_{max}}-3$ $\mathrm{dB}$), the quality of the sound gets worse and therefore, from the cabin we are listening to a dimmed sound of the flight crew voice.

Some of the newer aircraft PA systems are done by using DSP (Digital Signal Processing) systems, also within the limited range, but the quality of the sound is more improved.

References: Dimov Stojce Ilcev, "Global Aeronautical Communications, Navigation and Surveillance (CNS)", June, 2013.

• Looks like a nice answer but any chance you could add a source or reference to back this up? :) – Thunderstrike May 30 '15 at 16:06
• Are you sure about that ? You have any source ? Because what you describe is the standard "telephone" filtering (weather it is POTS or VoIp), and I don't see in any way how keeping a wider bandwith would interfere with radio comm. These are not at all the same freequencies! – kebs May 30 '15 at 16:12
• The RF spectrum starts at frequencies of 3 kHz I never heard of any radio communication device using frequencies below 100 kHz... moreover in airliners ! Remember that VHF airband is around 100 MHz and long-range HF is around 2MHz. – kebs May 30 '15 at 16:24
• This seems to be completely wrong. There is no way band-filtering sound can prevent interference with RF signals. The sound is transmitted over a narrow RF band anyway, the frequency distribution of the sound has nothing to do with the frequency of the RF signal. – ithisa May 30 '15 at 16:30
• To me this is more of a 'how' than a 'why'. Public address systems are well-understood technology and it shouldn't be difficult to build ones in planes that work well, even under adverse sound conditions. Such sound systems would, however, be more expensive. I strongly suspect that is the 'why'. – Carey Gregory May 30 '15 at 17:24

The microphones used in aviation headsets are simple electret microphones, with high vocal clarity but low response at high frequencies. They provide a quality similar to standard telephone service. Since the pilots use these microphones to speak on the internal PA, there's no reason to install a high fidelity PA fed by a low fidelity microphone.

The PA's purpose is clear transmission of information, same as the radio. High voice quality is a very low priority, and it's also difficult to control without a custom microphone setup for each individual pilot.

It's also worth noting that your basic assumption is that a higher fidelity system would improve the passengers' ability to understand the crew. This is not actually true. Low-bandwidth systems such as telephone operate on the principle that speech can be understood as well (and sometimes better) with its high frequency components removed. This offers the additional advantage of requiring lower channel capacity per user of the system (in both telephone and radio comms).

A high quality audio address system, such as that used by a radio DJ, requires a quiet environment, a high-quality, highly-sensitive microphone, and a speaker with a consistent voice, who maintains a consistent distance from the microphone. None of these are practical in an aircraft cockpit.

A slightly lower-level system, such as that used by a speaker on a stage, would use a microphone on an earpiece in a similar location to that of the pilot's microphone. However these are condenser microphones, which are non-directional and highly sensitive. Such a microphone would be overwhelmed by background noise in an aircraft cockpit.

In both the radio setting and that of the speaker on a stage, there is also a dedicated sound engineer who is monitoring the level of the speaker and making adjustments to the microphone output. Equalizers, compressors, and reverb are also employed among other effects to improve the quality of the speaker's voice.

• You don't need recording studio quality equipment to improve a PA system. PA systems on planes should equal or exceed the quality and intelligibility of PA systems frequently encountered in shopping areas, sports arenas, schools, etc. At the very least, a volume boost is needed. I've been on many flights where the PA was unintelligible because the volume was simply too low. – Carey Gregory Jun 1 '15 at 17:42
• Thanks for your answer. While I agree with most (not all) of your statements, I understand your answer as "can't really do better". I can't agree with that, some improvements -can/need to- be done. – kebs Jun 2 '15 at 9:37
• simple electret microphones, with [...] low response at high frequencies: This is simply not true, a 1\$ micro capsule has natively a very good bandwidth. But headset manufacturers do probably cut it down to avoid larsen effect. – kebs Jun 2 '15 at 9:55
• @kebs - I stand corrected on the microphone's frequency response. You are probably correct that they reduce it further down the line, though. – Joel M. Jun 2 '15 at 13:47
• Are you seriously telling me it's impractical to train flight crews to speak intelligibly on a PA? The same argument applied to radios would be met with laughter. And if someone with good hearing such as myself often finds PA volumes too low then I think it's a safe bet they often are too low, which shouldn't even be possible. I believe they most certainly can practically do better. – Carey Gregory Jun 2 '15 at 13:56