Intelligible audio is critical to the audience’s experience of any audio visual presentation. Spoken word and any programme audio must be clearly received by listeners, and the level and quality must be appropriate so listeners remain comfortable for the duration of each session.
Throughout the design process, attention must be paid to the different requirements of each output and key stakeholder group so the overall solution is adequate. For example:
Recording and conferencing systems may require a different mix of audio to any room PA system, and it may be delivered via a non-traditional interface such as USB
Hearing impaired listeners must be assisted to clearly discern the important audio over any background noise (minimum signal to noise ratio)
Any installed public address system must suit the physical acoustics of the room so listeners are not disadvantaged.
The skills required to commission an audio system are different to those required to program/configure the equipment, and technical managers should take care to ensure the build team includes an appropriate mix of ‘analogue’ and ‘digital’ skills and experience.
(Assistive) Listening System
Technology describes in the building code as being compliant for hearing augmentation. May be induction loops, infra-red or radio frequency.
The decibel is a relative unit of measurement used widely in acoustics. It is a logarithmic unit describing the ratio between a measured level and a reference value.
One common ratio to AV people is Sound Pressure Level (SPL).
Weightings are applied to sound pressure scales to reflect the performance of the human ear under certain conditions. Most typically encountered are:
The professional or technical specialist with responsibility for designing the electronic and electroacoustic (PA) systems and coordinating with other members of the design team. May be employed by the institution or an external consultant.
AV designer may also have responsibility for leading electroacoustic scope, where briefed.
Digital Signal Processor (DSP)
Any digital techniques for the manipulation of an electrical system.
In installed sound systems the term is often shorthand for the subset or equipment which performs audio acquisition, processing and routing.
Emergency Warning and Intercom Systems; one common form of warning system for fire and other emergency events.
The communication of information for people who are Deaf or hearing impaired by using a combination of audio, visual, and tactile means. (AS1428.5)
As well as assistive listening systems, can include warning lights, sign language and other non-AV provisions.
Speech intelligibility is a rating of the proportion of speech that is understood.
Signal to noise ratio
The difference between the measured sound level and the noise floor (all other sources). Expressed in dB.
The Speech Transmission Index (STI) is an objective metric ranging between 0 and 1 representing the transmission quality of speech with respect to intelligibility by a speech transmission channel
Fire detection, warning, control and intercom systems—System design, installation and commissioning
Part 4: Emergency warning and intercom systems
Design for access and mobility
Part 5: Communication for people who are deaf or hearing impaired
Audio Coverage Uniformity in Listener Areas
(Supersedes ANSI/InfoComm 1M:2009)
Design for Access and Mobility – Buildings and Associated Facilities
Fire Detection and Alarm Systems in Buildings
Audio/video, information and communication technology equipment – Safety requirements
National Building Codes
National Construction Code
The Building Code
Microphones may be fixed or portable (wireless). When deployed thoughtfully they help deliver clean, intelligible speech to the audience and other systems.
A fixed microphone is usually provided at any lectern and is typically of flexible ‘gooseneck’ construction for ease of adjustment. Microphones employing a cardioid pickup pattern help limit unwanted audio and maintain a good signal to noise ratio. Typical fixed microphones still provide a balanced analogue output for simplicity and compatibility.
Wireless microphones are used in most theatres and larger flat-floor spaces to provide freedom of movement to presenters. Balanced analogue audio is still a common output, with a network connection increasingly delivering multi-channel audio to and from professional equipment, as well as providing system event monitoring
Handheld wireless microphones are preferred by some presenters and generally employ a cardioid pickup pattern
Lapel microphones leave presenters with both hands free. Most presenters aren’t skilled in microphone technique and may not position the lapel mic well, so an omnidirectional pickup pattern is usually more appropriate.
Headset microphones position the microphone capsule at a fixed distance from the presenter’s mouth and offer the best chance of clear speech pickup. As input sensitivity is comparatively lower than for an omnidirectional lapel, the result is typically a high signal to noise ratio and better gain before feedback. Some users dislike headsets and they may not be appropriate for use in all common learning spaces but provide excellent utility in spaces such as wet labs.
Specialist microphones are increasingly used for conferencing and collaboration systems and may be ceiling, wall or desk mounted.
Designers should be careful when using boundary microphones. As they are deployed as a ‘catch all’ solution, background noise is acquired at a higher level and may provide unwanted masking in assistive listening systems.
The requirement for discrete, installed source devices in typical learning and meeting spaces has lessened with the increasing reliance on user-provided devices (BYODs) and network-delivered content. Typical provisions include:
‘Resident’ PC and provision for connection of BYODs via commonly-supported cable types.
A network presentation gateway which allows network connection of BYODs to presentation systems
Dedicated AV decoders for the receiving of network video streams
IPTV, MATV or other content appliances
Audio is almost universally embedded in digital video streams from these devices, but provision for analogue audio input may be appropriate in some spaces. Consider the functionality required, and what analogue inputs need to be afforded (e.g. 3.5mm stereo jack, XLR mic or line input(s) for an event mixer)
Videoconferencing, capture and collaboration systems are regularly encountered and may deliver two-way audio via analogue, embedded digital or USB.
Institutions increasingly rely on the use of digital signal processing (DSP) techniques to handle acquisition, mixing and processing and maintain high signal to noise ratios throughout. Audio DSPs can be tightly integrated with room control systems and lessen the potential for well-intentioned tampering.
Processor size and functionality will be determined by the required number and type of inputs/outputs and the complexity of the planned activities. Modern AV switchers/routers often include an audio DSP which may provide adequate flexibility and functionality.
Whilst DSP resources have traditionally been deployed as a part of each local AV system, it is now viable to pool DSP resources at a building or campus level. In addition to functional requirements, institutions must consider the needs of those programming and supporting the systems when determining the most appropriate topology.
Some organisations will find benefit in deploying a dedicated DSP in a particular space or floor. Others may be looking to leverage centralised DSP deployment at a building or campus level. The appropriate solution should be based on the functionality required for all users, including those supporting the systems.
The selection of audio DSPs within an organisation should consider the needs of the support staff who will manage them remotely. Where practicable, the number of DSP types should be reduced to ensure appropriate staff are trained on the operation of configuration and management software. Once an institution has selected their DSP platform, they should consider training key technical staff. Online training is provided by many manufacturers, and those intending to develop the configuration and programming may benefit from additional classroom training.
Designers should leverage the available diagnostic tools provided in DSPs. Signal flow tracing, test tools, metering and remote audio monitoring allow technical staff the ability to pinpoint specific problem areas, and enable them to deploy and manage incident response more effectively.
For those institutions deploying rooms built upon standardised designs, a common approach is to develop a standard “site file” with a typical layout, naming, I/O configuration and testing tools.
Institutions in Australia and New Zealand benefit from good access to manufacturers and distributors, and may choose to seek their advice on development of these site files. This approach may help ensure a good balance of standardisation and the flexibility needed to cover most scenarios.
Site files should be provided before the build commences, so any issues can be identified prior to final system commissioning by the AV integrator or internal staff.
Audio mixing and processing equipment generally:
Must be able to be reliably interfaced to the room control system
Must accept input audio in all required formats required in the space e.g.:
Line- and microphone level analogue audio with phantom power for condenser mics
Multi-channel digital audio via a network stream or other industry-standard interface (e.g. USB)
May include one or more de-embedders to extract audio from an HDMI stream for downstream mixing and processing
Must provide discrete signals to all output and capture devices; typically analogue audio at line level or digital audio in the institution’s preferred format
Must include all mixing, routing, dynamics and monitoring functionality to support the planned use cases
Fixed-architecture DSPs provide defined signal paths and routing with limited processing
Open-architecture DSPs require skilled specialists to define the internal audio architecture using a library of configurable components. They generally allow the creation of larger, more complex systemsWhere video- or teleconferencing is integrated with room AV, Acoustic Echo Cancellation (AEC) must be employed to prevent undesirable effects caused when incoming audio is picked up by microphones and retransmitted.
AEC may be achieved in hardware (DSP) or software (many ‘soft’ VC codecs), but only one AEC instance should occur. If software AEC cannot be disabled, additional hardware processing may be detrimental
In large spaces, it may be appropriate to define multiple AEC zones, such that microphones are referenced to their local loudspeakers
In all cases, the configuration and commissioning of audio processing and mixing systems must be undertaken by staff with appropriate direct experience in assembling audio systems and tuning them to the physical environment. PA system commissioning is one area where the requisite skills are largely the same as they were before the advent of digital processing.
Audio systems feed a number of discrete outputs including:
Speech and programme loudspeaker systems may be discrete or combined as a single system
Programme systems are typically configured as stereo, however a mono arrangement is more appropriate for some layouts and multi-channel systems may be required in those labs and theatres that require discrete audio zones/channels or immersive audio.
Voice systems are generally mono, but multiple channels may be provided so each can be delayed slightly to improve intelligibility over distance.
Recording and conferencing:
One or more audio feeds are required to each system for capture of intelligible audio
Programme capture may be stereo or mono, depending on the capture device
Speech may feed a separate input or be mixed with programme audio. In the latter case, ensure speech is clearly audible above programme. Ducking can be utilised if required, and should be implemented sensitively to decrease a distracting ‘pumping’ effect caused by too-rapid release.
Assistive listening systems:
A mono sum of speech and programme is provided to most assistive listening systems, ensuring speech is clearly audible above programme
Speech must be derived from those individual microphones actually in use and should specifically exclude audio from room effect/boundary mics which can dramatically decrease signal to noise ratio and decrease intelligibility and amenity to the hearing impaired.
An effective electroacoustic design is essential to assist presenters in communicating spoken word and other programme material to the audience.
Many institutions will develop a preference for one or more narrow ranges of speakers based on performance, voicing and reliability, but the final choice of speakers should always be reviewed for suitability in each space.
Design of the speaker system will be informed by the need to achieve:
Effective coverage of the listening area(s)
Acceptable uniformity of SPL throughout the listening area
Good intelligibility and high gain before feedback for amplified voice; and
Reproduction of programme audio appropriate to the identified use cases
Typical performance criteria for a practical sound reinforcement system are
SPL of 65dB(A)
Intelligibility in the ‘excellent’ (0.75+) range (STI per IEC60268:16)
SPL of 85dB(A)
Uniformity (AVIXA A102.01:2017 Audio Coverage Uniformity in Listener Areas)
Coverage envelope of 6dB
Practical design strategy
Amplified voice is not expected in most small spaces, so amplification is typically limited to programme sound.
In small rooms, this is usually from a pair of speakers at the front or the room, or a sound bar where flat panel displays are used. In flexible/collaborative spaces with multiple displays or no single teaching space an array of ceiling speakers may be more appropriate.
Where microphones are used, acoustic feedback is an undesirable side effect which must be prevented.
‘Front’ programme speakers are usually augmented or replaced with ceiling speakers to provide more uniform distribution of amplified voice. Uniformity allows technical staff to better commission the sound system to provide adequate SPL and high gain before feedback
The aim of any video conference space is to provide as natural a conversational environment as is practical. Depending on room size, programme and ‘far end’ sound are reproduced through a combination of
In larger conferencing spaces, ceiling speakers may be wired in multiple circuits to reflect the number and placement of microphones and improve Acoustic Echo Cancellation (AEC) processing
Larger spaces require a bespoke approach which responds to the physical acoustics and best supports identified uses.
The audio feeding these speakers should be delayed so the sound arrives slightly later than that from the front, ensuring listeners still perceive sound as localised to the front of the room.
Care must be taken in the design to use speakers appropriate for the task, and to place them to best effect. The manufacturer’s datasheet is informative but must be read closely so calculations are based on realistic data. Designers should particularly review the following items and inspect polar plots or supplementary data to reach a decision.
Generally expressed as xxdB 1W/1m, or “What SPL will be measured on-axis at 1m from the speaker for a 1W input”
The vertical and horizontal distribution at a nominated frequency (normally 1-2kHz – in the voice range). Lower frequencies tend to be omnidirectional and dispersion narrows with increasing frequency
Many professional ‘box’ speakers use a rotatable waveguide on the high frequency element to permit vertical or horizontal mounting.
Line arrays tend to have wide horizontal dispersion (above 150°) but very narrow vertical (10-20°)
Ceiling speaker dispersion depends on construction, and commonly ranges between 90 and 150°.
Ceiling speakers with narrow dispersion should be considered in spaces with specular walls/façade to focus energy downwards
Speakers are ordered for direct, low impedance connection (4-16Ω) or use on a high impedance line at 70/100V.
Speakers rated for 100V will work fine on a 70V amplifier, though all power tappings are halved. Many commercial speakers are labelled with both 70V and 100V power ratings
A 70V speaker cannot always be assumed suitable for use on a 100V line – designers should check before specifying
The frequency response of a speaker is expressed as the range of frequencies able to be reproduced e.g.: 85Hz-15kHz
On its own, this frequency range provides an incomplete picture – we also need to know the tolerance applied.
This is often the point 6dB below the on-axis reference level, but some manufacturers measure at the -10dB points
Cinema and other more critical speakers may be described with a tolerance of ±3dB of mean
Loudspeaker polar plots are very informative if the printed data is not explicit. These present a graph of level between 0° and ±90°, helping designers understand the on- and off-axis frequency response of each particular unit for each of the key frequencies.
The electroacoustic implementation for any space includes design and configuration of active systems for the amplified reproduction of audio. Particularly in cinemas, performance venues and similarly critical spaces the required electroacoustic performance may heavily influence the physical acoustics.
In particular, we are seeking to
Provide audible speech and programme sound to all listeners (SPL, uniformity)
Maximise speech intelligibility (STI)
Achieve an acceptable frequency response by control any peaks or troughs that might cause discomfort
The designer will apply their professional judgement to shortlist appropriate components for the sound system based on the physical environments, the loudspeaker’s technical parameters and the designer’s subjective opinion on its voicing. An institution’s equipment or supplier preferences will influence shortlisting, but designers must always consider what is most appropriate for each space.
In larger spaces and those able to be linked or divided, speakers must be logically zoned. The simplest zoning may be for adjacent, linkable spaces, though in large spaces the audio feeding each zone may be delayed slightly to reduce late arrivals and improve intelligibility. In complex spaces, an audio DSP affords designers the capability to change level, equalisation, delay and dynamics instantaneously to cater for each possible room configuration.
Industry-standard tools and calculation methods should always be employed to refine each design, including:
Nomination of safe sound pressure levels, especially for extended listening
Calculation of the electrical power required (EPR) by the selected speakers to achieve the performance target
Calculation of the room’s stability (resistance to feedback) based on the needed acoustic gain (NAG), the potential acoustic gain (PAG) and the number of microphones required
Spacing of ceiling loudspeakers to achieve the desired uniformity based on their dispersion and sensitivity and the height of the ceilings
Calculation of audio delays to each zone to better synchronise arrivals or appropriately time-align different components
It may be appropriate to commission a detailed electroacoustic model to predict the coverage (SPL and uniformity) of amplified systems and/or unamplified speech in certain spaces, and to predict the intelligibility that may be achieved with the nominated system, room geometry and intended surface treatments.
These models are prepared by specialists using an electroacoustic simulation program. A number of programs are available, and each targets a different market.
Institutions will most commonly encounter AFMG’s Electro Acoustic Simulation for Engineers (EASE) as it is accessible to – and can be afforded by – the majority of professional audio contractors. Most manufacturers release EASE data for their loudspeakers (including line arrays) so they can be accurately modelled. Institutions should consider the availability of an infra-red module for EASE which will also permit modelling of some infra-red assistive listening systems to increase confidence that the system will meet Code.
Acousticians and sound system designers working on more complex spaces may prefer a specialist program such as Odeon which provides more comprehensive tools for acoustic analysis. Odeon also has the capability to model a wide range of speakers, though not all manufacturers can provide their data in the required common loudspeaker format (CLF).
Many specialist and DSP-steerable loudspeakers can only be modelled in proprietary tools provided by the manufacturer.
Members should remember that any model is used to validate and fine-tune the design. First principles calculations and the designer’s experience must still inform the type, location and power requirements of speakers.
The level of detail in the model and the accuracy of surface material parameters directly affect the accuracy of the prediction, and the designer and modeller should agree the appropriate level of detail for each project.
A significant proportion of the population is affected by some form of hearing impairment, and hearing augmentation systems are deployed to provide equitable access. Whilst the term itself is broad and includes closed captioning and signing, in this context we consider the provision of assistive listening systems only.
As the basic requirement is always to provide equitable access, the goal must be to provide clear, measurable amenity to the affected group. Legislation and the relevant National building codes define a minimum provision and institutions may choose to offer a higher level of coverage. Building code requirements differ between Australia and New Zealand.
Three types of systems generally satisfy the Code requirement in most countries:
Audio Frequency Induction Loop Systems (AFILS)
Current is passed through a metal loop in the floor (typical) or ceiling and generates a magnetic field. This is converted back into electrical energy via a telecoil in the hearing aid.
Great majority of adult hearing aids in Australia and New Zealand include a telecoil.
Hearing aids for children and youth must be replaced more frequently and telecoils are often omitted
Hearing aids in other countries may not be telecoil-equipped in the same proportion as Australia and New Zealand
AFILS-equipped rooms cannot be considered secure/private as the magnetic field can propagate far outside their boundaries.
Systems in adjacent rooms may interfere with each other.
Infra-Red (IR) systems
Audio is modulated onto an infra-red (light) carrier and demodulated by a personal receiver. Headphones or a neck loop (to drive a telecoil-equipped hearing aid) are used.
IR systems are easier to maintain and can provide confidentiality as the light is blocked by solid walls and heavy curtains. Conversely, IR systems may be adversely affected by sunlight.
Room surfaces may reflect some IR energy and provide a small coverage improvement which assists in meeting code.
IR systems require near line-of-sight to operate, meaning the receiver is typically visible on or near the listener..
Radio Frequency (RF) Systems
Audio is modulated onto an RF carrier in the ‘low interference potential’ spectrum and received by a personal receiver. Headphones or a neck loop (interface to a telecoil-equipped hearing aid) are used.
Digital RF modulation techniques are now common and may allow higher density of systems in learning environments
WiFi streaming systems may satisfy code where performance is demonstrated to comply and will increasingly allow users to use an app on their personal device Australia: AS1428.5 clause 4.5.1: Specialised systems using advanced or other communication technologies such as digital communication systems… New Zealand: There is no clear ‘deemed to satisfy’ provision for a WiFi-based RF system. Technical Manager should pursue this with the project Access Consultant and Building Surveyor for consideration as an alternative solution.
Sound Field Amplification Systems (Soundfield) are encountered in some environments, and AS1428.5 acknowledges they may provide some benefit for mildly impaired listeners. The standard does not, however allow Soundfield systems to be used as compliant Assistive Listening Systems..
Signage, generally using the International Symbol for Deafness, must be installed as part of any hearing augmentation system, and must be designed and installed according to local building codes.
Many institutions have developed internal operational policies which govern the deployment and maintenance of hearing augmentation systems to best address the requirements of their own populations and facilities. They are intended to be applied uniformly throughout the institution, so students are well supported, and are often developed by – or in close cooperation with – the institution’s Accessibility or Disability Support teams.
The implementation and performance of any such system must always meet or exceed Australian or New Zealand code requirements, so technical managers should coordinate closely with these teams to ensure these requirements are met.
Technical Managers must take care with any new building or substantial refurbishment which will require certification against the prevailing Building Code or standards by an independent building surveyor. The operational policy must be reviewed for suitability by the project’s Access Consultant and approved by the Building Certifier/Surveyor before a Certificate of Occupancy will be issued. Where an institution’s operational policy differs from the code it must be assessed as an ‘alternative solution’.
The requirement for assistive listening systems is defined in National Construction Code (NCC) clause D3.7:
Minimum compliant coverage requirement is 80% floor area (AFILS) or 95% (IR, RF).
Whilst the NCC is silent as to the underlying standard, AS1428.5 is the only Australian standard dealing specifically with the design, commissioning and certification of hearing augmentation systems. This standard expands on the minimum performance requirements for AFILS described in AS60118-4 and adds performance criteria for IR and RF systems. Prescriptive testing and certification requirements are also defined for each system.
Signage is mandated in the NCC at D3.6 and defined in AS1428.1. Designers should exercise caution as international signage and that provided by manufacturers often does not satisfy Australian standards:
Only a single colour blue is compliant
The International Symbol for Deafness is trademarked locally, and must be used as drawn
The location of signs and form of words must be compliant
Braille may be required on some signs
The Deafness Forum (trademark owner) publishes a signage guide which may aid designers to develop compliant signage.
Note: The NCC defines different criteria (clause H2.13) for Class 9b and 10 public transport buildings only. Designers should be careful if designing systems for public transport interchanges on campus.
The New Zealand Building Code is less prescriptive than Australia’s, however several provisions apply to communal non-residential and commercial buildings:
The underlying standard (NZS4121) provides guidance to those responsible for design to enable people with disabilities to use the building with the same convenience as those who do not have disabilities. Appendix H provides the framework for ‘listening systems’; including:
Usable by people with and without hearing aids (such aids with or without telecoils)
Suitable for the intended use
Not to cause or be subject to interference
Should be infra-red where privacy is required
Field strength to AS1088.4 (now AS60118.4)
No defined minimum coverage requirement
Noted as inappropriate where crosstalk/spill will occur between adjacent spaces
No additional performance/coverage criteria are defined
Radio Frequency (H3.3)
Permitted on ITU-allocated VHF frequencies
Permitted, but no performance/coverage criteria are defined
Induction field radio system (H3.4)
Permitted on 3175kHz allocated by ITU
No additional performance/coverage criteria are defined
Appears intended primarily for ‘tour guide’ systems
The signage requirement is defined in E3 and provides only broad guidance:
The International Symbol for Deafness must be used and rendered proportionally as shown in the standard.
The symbol shall be white over a “safety blue” (BS5252 colour 18E53) – though other colours may be used to fit the décor sufficient to adequate contrast.
The national building and fire codes in Australia and New Zealand require that emergency and evacuation messaging can be clearly heard by the population of a building.
Whilst other forms of alerting system will apply to some projects, it is most common to encounter an EWIS (Emergency Warning and Intercommunication System). Both Australia and New Zealand reference AS1670.4; the deviations identified in NZS 4512 concern the Fire trade, not AV.
This is a life safety system, so:
It is critical for designers to liaise closely with each project’s fire engineers to understand, allow for and implement strategies which ensure the building can be certified. The approach taken on any previous project cannot be assumed to be automatically applicable.
We must appreciate our sound systems are competing sound systems under AS1670.4 and may need to be shut down when any emergency warning system is broadcast.
If a combined EWIS/general PA is considered, AS1670.4:2015 requires loudspeakers to comply with AS7240.24. At the time of publication, only those speakers marketed specifically for EWIS comply, though most speaker manufacturers were evaluating certification requirements.
While this document cannot predict any certifier or fire engineer’s opinion, our ‘competing’ PA systems will often be evaluated as compliant on the following basis:
Auditoria, lecture theatres, large spaces (average SPL >75dB(A) is likely)
Fire contractor provides a failsafe EWIS mute line from their system
When an alarm is triggered, or the line breaks the trigger changes state.
This trigger may be dry (relay) or wet (voltage) depending on local circumstances.
AV system mutes audio on change of state.
Occasionally we are required to fully shutdown the amplifiers
Meeting and smaller rooms (average SPL≤75dB(A) is likely)
System is configured so average SPL is well below the EWIS – generally at least 10dB