Designing for Intelligibility & Energy Efficiency

© 2013 Eaton, All Rights Reserved.

Intelligibility vs. Audibility

© 2012 Eaton Corporation. All rights reserved.

Audibility & Intelligibility

• NOT the same!

• Think of the sayings:

• “I hear you LOUD & CLEAR”
• “I have you 5 by 5”
• Military/radio jargon for subjective loudness & clarity
measurements. 5x1 is loud, but not clear, 1x5 is not loud
but crystal clear

© 2013 Eaton, All Rights Reserved. 3

Audibility Defined

• Measured in decibels (dB, dBA)

• The state or quality of being perceptible
by the human ear
• Normal Calculations for Fire Alarm
professionals – fairly straight-forward –
15 dBA over ambient
• Double the distance from the source &
lose 6 dB – examples:
• 100 dB at 10’, 94 dB at 20’, 88 dB at
40’
• Measurable with Type II Sound Level
Meter
• Goldline, etc.

© 2013 Eaton, All Rights Reserved. 4

Intelligibility Defined
• “is a measureable aspect of
electronic voice transmission
systems that indicates the degree
that human listeners will be able to
understand the voice messages
transmitted through them.”
• Fire Protection Research
Foundation (FPRF)
• Measurement is usually expressed
as a percentage of a message that
is understood correctly
• A variety of test methods are
available to measure intelligibility

© 2013 Eaton, All Rights Reserved. 5

Latest Intelligibility Standard Overview

• Mandatory for DoD MNS – UFC

• New UFC requires exterior testing
• Computer designs – EASE Evac
UNIFIED FACILITIES CRITERIA (UFC)
• NFPA 72 2013
• Clarified intelligibility testing requirements
• Up to governing laws, AHJ, or codes to
specify where voice signaling must be
intelligible
• In absence of codes, the system designer
must designate which ADS will have
intelligible voice EASE Evac -Speech Transmission Index (STI)

• Where required by AHJ or codes,

ADS/intelligibility plans must be
submitted for approval

© 2013 Eaton, All Rights Reserved. 6

Introduction to Intelligibility

• Whole new animal

• Requires knowledge of sound
• Design usually done with
computer modeling
• Renkus-Heinz
• Bose
• Others
• Very dependent on area
acoustics

© 2013 Eaton, All Rights Reserved. 7

Room Finish Affects Intelligibility

• Changing room finish from carpet

to ceramic tile will usually make
intelligibility WORSE, but may
make audibility better.
• Changing room finish from tile to
shag carpeting will usually make
intelligibility BETTER, but will
make audibility worse.
• Have to keep up with room finish
changes on projects.

© 2013 Eaton, All Rights Reserved. 8

Speech Intelligibility Test Methods
Subject-based – ANSI S3.2
• Phonetically Balanced Word Scores (PB) = (256 words
or 1000 words)
• Modified Rhyme Test (MRT)
Instrument-based – ANSI S3.5
• Articulation Index (AI), now referred to as Speech
Intelligibility Index (SII)
• Calculated from acoustical measurements of speech and
noise
• Speech Transmission Index (STI)
• Predicts how the characteristics of the transmission channel
affect speech intelligibility
• Speech Transmission Index for Public Address (STIPA)
• Modified method of STI
• Common Intelligibility Scale (CIS)
• Developed to show the relationship between the different
test methods

© 2013 Eaton, All Rights Reserved. 9

Intelligibility Measurement Standards

• Designing for intelligibility takes practice

• Design for COMPLETED & OCCUPIED spaces unless specifically directed
to do otherwise
• Measured in Common Intelligibility Scale (CIS)
• NPFA - 0.7 or higher usually passing
• Trend is to measure in Sound Transmission Index Public Address (STIPA)
• NFPA - 0.5 or higher usually passing
• DoD facilities are required to test voice intelligibility using CIS or STI
• Army & Air Force – Minimum CIS 0.8 (0.75 can be rounded to 0.8)
• Navy & Marines – Minimum CIS 0.7
• MAKE SURE TO KNOW WHAT YOU ARE MEASURING!

0 STI 0.3 0.45 0.6 0.75 1.0

Bad Poor Fair Good Excellent

0 CIS 0.48 0.65 0.78 0.88 1.0

© 2013 Eaton, All Rights Reserved. 10

Design General Rules Of Thumb

• LOTS of speakers at lower wattage taps

• Can be as high as 8 times the speakers
used to achieve audibility alone
• Example - Airports
• Increasing the wattage of speaker often
distorts the message
• Ideal point-source (speaker to listener)
is 20’ or less
• Make sure all speakers in aural range
are IN PHASE
• Ceiling height is speaker spacing
• Refer to NFPA documents for guidance

Important: Speaker placement, type of

speaker, spacing and power

© 2013 Eaton, All Rights Reserved. 11

Frequency Response Range

• Humans hear by frequency response

• Most energy of normal speech is 500 Hz
to 4 KHz
• Fire Alarm/MNS systems generally
produce frequency responses of 400 Hz
to 4 KHz
• Commercial sound systems generally
produce frequency responses of 125 Hz
to 12.5 KHz.
• “Pro” sound system generally produce
frequency responses of 20 Hz to 20 KHz.
• Humans can hear from around 50Hz up
to 22kHz and most past 50 cannot hear
above 10 KHz

© 2013 Eaton, All Rights Reserved. 12

Frequency Response
The wider the frequency response of a speaker, the better it
reproduces the frequencies in the original signal
• If the loudspeaker reproducing speech can’t handle
certain frequencies in that speech, you’ll lose some of
the information in what was originally said.
• Wider frequency response means that the speaker
makes the speech easier to understand to whomever
is listening
• The most critical elements of speech, the consonants,
lie above it 3 kHz.
• Frequency that distinguishes the "s" in "sailing"
from the "f" in "failing" occurs between 4 kHz and
14 kHz.
• Emergency – Reduce the chances of Stair B
being misinterpreted as Stair D, Stair C or Stair E

Wider frequency response = Higher intelligibility

© 2013 Eaton, All Rights Reserved. 13

STIPA & MNEC/FA

• STIPA uses frequencies of 125 Hz to 8 KHz

(Typical MNS/FA systems produce 400 Hz to
4KHz)
• Best case testing of many MNS/FA systems
with no appreciable wiring and near perfect
acoustics in the aural field is 0.75 STIPA (0.5
passing) or 0.85 CIS (0.7 passing). This
means that acoustics and speaker
performance must be evaluated carefully
before installing a system.

© 2013 Eaton, All Rights Reserved. 14

Cautions For Intelligible Design

• Not all areas can achieve

intelligibility
• Think of historic building,
polished marble floors & walls,
hard plaster ceiling, no
carpeting, no baffling
• Machinery & high-ambient SPL
areas may not be able to achieve
good intelligibility
• Reasonable approach for such
areas may be rally points where
intelligibility can be obtained
(people hear audible signal &
have to go to a place to
comprehend an intelligible
signal)

© 2013 Eaton, All Rights Reserved. 15

Designing for Outdoors

Additional parameters
• Wind
• Temperature
• Humidity
• Changing of the environment due to
seasons, age, and time
• Vegetation growth affects sound
• Designing a Wide-Area MNS in the
middle of winter when trees are bare
without considering trees in full bloom
can greatly affect the intelligibility of the
system Same principals for indoors apply to
outdoor areas

© 2013 Eaton, All Rights Reserved. 16

Intelligibility Testing

Before testing, make sure:

• Audibility is achieved
• Audio system is balanced
• Amps are calibrated
• Control room is QUIET
• Radios, talking, door slams, etc. will mess up testing
• Construction is done

© 2013 Eaton, All Rights Reserved. 17

NFPA on Intelligibility

© 2012 Eaton Corporation. All rights reserved.

History of Intelligibility – NFPA 72
1993 Edition
• 1st edition having mandatory requirements for notification appliances
• Textual audible appliances (speakers) required to reproduce normal voice
frequencies
1999 Edition
• Included definition of Voice Intelligibility
• “Audible voice information that is distinguishable and understandable”
• Emergency systems have messages with voice intelligibility - most information
was placed in the annex.
• There was a proposal to add a requirement to the body of the code that
systems provide .7 CIS. This was rejected by membership
• References to IEC standards for quantitative measurement approaches
2002 Edition
• Added Guidance to designers
2007 Edition
• Intelligibility referenced in more locations
• Incorporated Annex E on MNS, which generated renewed focus on intelligibility

© 2013 Eaton, All Rights Reserved. 19

History of Intelligibility – NFPA 72

• Task Group formed at end of 2007

Code cycle to develop an approach for
mandating quantifiable measurement
of voice intelligibility
• Task recommended adoption of ISO
7240-19, Design, installation,
commissioning and service of sound
systems for emergency purposes
• FPRF project formed after ROP
• New Definition for Intelligible
• New Definition for Intelligibility
• New Definition for Acoustically
Distinguishable Space
• Expanded Annex to provide guidance
on design for intelligibility and methods
for quantitative measurement

© 2013 Eaton, All Rights Reserved. 20

For Intelligible Design

Will need to know:

• All room dimensions
• Room use
• Room occupancy
• Room finishes &
treatments
• Speaker polar plots

“Intelligibility of Fire Alarm and

Emergency Communication Systems”
Fire Protection Research Foundation
November 2008

© 2013 Eaton, All Rights Reserved. 21

Intelligibility & ADS Defined – Chapter 3
NFPA 72 2010 & 2013: New/Revised Definitions

• Intelligibility
• The quality or condition of being intelligible
• Intelligible
• Capable of being understood; comprehensible; clear
• Voice communications used in emergency communication systems (ECS)
• Change from 2007 – Audible voice information that is distinguishable and
understandable
• Acoustically Distinguishable Space (ADS)
• Emergency Communication System notification zone, subdivision or
physically defined space
• Distinguished from other spaces because of different acoustical,
environmental or use characteristics (reverberation time, ambient sound
pressure)
• All parts of a building intended to have occupant notification are subdivided
into ADSs

© 2013 Eaton, All Rights Reserved. 22

Documentation - Chapter 7
NEW to NFPA 72 2013
• Minimum protocol for documentation; May or may not be required by AHJ
• Some information was dispersed among a number of chapters.
• NEW - Written narrative must be included as part of the submittal drawings
• Forms completely revised so they are easier to use – with a basic form for straight
forward systems and supplemental forms for more complex systems
• Directives to provide
• Ambient noise and audible design sound pressure levels
• Extensive emergency communication systems design document
requirements
1. Emergency Communication System Supplementary Record of Completion
2. Mass Notification System Supplementary Record of Inspection and Testing
• Includes Sound pressure levels and System intelligibility CSI or STI with
locations, values and weather conditions.

© 2013 Eaton, All Rights Reserved. 23

Voice Intelligibility & ADS - Chapter 18
NFPA 72 2010 & 2013: New/Revised Definitions
• Voice Intelligibility within the ADS
• Where voice intelligibility is required, voice communications systems shall
reproduce pre-recorded, synthesized or live messages with voice intelligibility
• ADS determined by system designer during planning and design of all ECS
• Each ADS shall be identified as requiring or not requiring voice intelligibility
• Update in 2013
• Requirement to document which locations will have audible notification & which
areas will not
• Statement that coverage area for audible occupant notification shall be as required
by other governing laws, codes, or standards such as local building or fire codes.
• Requirement to document sound pressure levels of the notification appliances for
the various coverage areas for use during acceptance testing.
• If required by AHJ, documentation may need to be submitted for review and
approval.
• Intelligibility shall not be required to be determined through quantitative
measurements
• Quantitative measurements as described in Annex D shall be permitted but not
required
© 2013 Eaton, All Rights Reserved. 24
Steps to Identifying ADS

Each ADS shall be identified as requiring

or not requiring voice intelligibility

© 2013 Eaton, All Rights Reserved. 25

Indoor Voice NOT Required

• Moved from Chapter 24 to Chapter 18

• Unless specifically required by other governing
laws, codes or standards, or by other parts of
this Code, intelligibility shall not be required in
all ADS. For example:
• Private bathrooms, showers rooms, saunas
etc.
• Mechanical/electrical/elevator equipment
rooms
• Elevator cars
• Individual offices
• Kitchens
• Storage rooms
• Closets
• Rooms/areas where intelligibility cannot
reasonably be predicted

© 2013 Eaton, All Rights Reserved. 26

ECS/MNS – Chapter 24
NFPA 72 2013
• Numerous changes to correct minor errors and references
• Improvements on ECS operations, intelligibility and documentation
• Sound and communications issues continue to evolve
• NEW
• Microphone Use
• Users shall be provided with posted instructions for microphone use
• Characteristics of the system microphone are important ergonomic factors that
affect Voice Intelligibility
• Message Templates
• Introduced into the Code, which requires that messages be developed for each
scenario in the emergency response plan
• For all evacuation messages, a tone in accordance with Chapter 18 must be used with
2 cycles before and after the message as a minimum
• All test messages must clearly state: “This is a Test Message.”

© 2013 Eaton, All Rights Reserved. 27

Voice Evacuation in Sleeping Rooms
Chapters1818&&2424– -NFPA
Chapters NFPA72
722013
2013
• Low Frequency Sounders (18)
• Effective Jan 1, 2014, audible appliances provided for sleeping areas to awaken
occupants shall produce a low frequency alarm signal that complies with
• Alarm signal shall be a square wave or provide equivalent awakening ability
• Wave shall have a fundamental frequency of 520 Hz + 10 percent
• Low frequency audible signal required in areas intended for sleeping and in areas
that might reasonably be used for sleeping i.e.
• Apartment bedroom and the living room area of an apartment as it might have
sleeping occupants
• Hotel guest rooms
• Voice Evacuation (24)
• In occupancies were sleeping accommodations are provided and the voice
message is intended to communicate information to those who could be
asleep, a low-frequency tone that complies with 18.4.5 shall be used for the 2
cycles before and after the message

© 2013 Eaton, All Rights Reserved. 28

Speech Intelligibility – Annex D
NFPA 72 2010 & 2013
Guidance on planning, design and testing of voice communication systems
• Acoustic performance parameters of chosen loudspeakers and placement are key to
determining how many appliances are necessary for adequate intelligibility
• Numerical count of appliances for a given design and protected space can NOT by itself be used to
determine the adequacy of the design
• Acoustical problems of certain placement constraints can be overcome through careful selection of
loudspeakers with requisite performance characteristics, rather than increasing their number
• Environmental Factors
• Ambient noise level in ADS
• Physical room characteristics
• Frequency and level of ambient sound pressure level might require components that have
a wider frequency range
• Fire alarm speakers designed in accordance with UL 1480 are only tested for and
required to produce frequencies of 400 to 4000 Hz (average range for human ears)
• Speech includes a wider range of frequencies
• STI and STIPA range from 125 Hz to 8000 Hz
• Under certain acoustic conditions, systems that do not produce the highs and lows, can
produce speech intelligibility that is less than desired
• Wider frequency response sounds better and is more intelligible to listeners
© 2013 Eaton, All Rights Reserved. 29
Voice Intelligibility for ECS – Supplement 2
NFPA 72 2010 & 2013
A well-spoken intelligent message in the listener’s native language can be
misunderstood if it is:
1. NOT audible to the listener
2. Distorted by the delivery system
3. Distorted by the acoustic environment
Factors affecting intelligibility
• Signal to noise ratio – Audibility is relative to the background noise
• Frequency response
• Harmonic Distortion – Reduction of clarity by distortion can be caused by:
1. Amplitude distortion caused by electronics/hardware
2. Frequency distortion caused by electronics/hardware or the
acoustic environment
3. Time domain distortion due to reflection and reverberation in the
acoustic environment
Noise, Frequency Response, Distortion &
Physical Characteristics Affect Intelligibility

© 2013 Eaton, All Rights Reserved. 30

 Voice Signal Path

Voice Signal Path

Talker Microphone Mixer Amplifier Speaker Rooms/Space Listener

Language Language
Speed Bandwidth Distortion Noise, Reverberation Hearing
Articulation

Assumed Normal Intelligibility Measures Assumed Normal

© 2013 Eaton, All Rights Reserved. 31

NIST – Guidance on
Emergency
Communications

© 2012 Eaton Corporation. All rights reserved.

Message Content & Delivery
NIST Technical Note 1779 – Feb. 2013
• NIST workshop identified a need for further research on
messaging and communications strategies
• Guidance on how to create and disseminate emergency
information (audible and visual)
• Understanding of human response to emergency
communications
• Ways to improve response via more effective messages,
message formatting and message dissemination
• Guidance on both alert signals and warning messages
• Latest Technical Note Draft – Nov. 2013
• Message templates provided/Guidance on creating messages
• Based on building emergency type and technology used
to disseminate the message
• Audible, text & even twitter examples
• Guidance on Emergency Message Testing
• The effectiveness of the emergency message
• Usability, Paraphrase, Readability

© 2013 Eaton, All Rights Reserved. 33

Barriers to Effective Emergency Response

• Perception
• Occupants need to perceive the warning
before they take action
• Factors that inhibit this process:
• Hearing impairments
• Visual impairments
• Situational conditions (sleeping, drug/alcohol
impaired)

• Attention
• Building may be equipped with audible and
visual distractions
• Comprehension
• Understanding the information provided by the
Emergency communications system
• Belief in Information
• Personalization of Information
• Protective Action

© 2013 Eaton, All Rights Reserved. 34

Emergency Communication Strategies
Alerts
• Imperative to disseminate an alert to let building
occupants know that a warning message will follow
• Audible – Sounds or series of sounds or words or
series of words
Warning
• Meant to provide information to building occupants
on the state of the emergency and what they are
supposed to do in response to that emergency
• Message Content should be specific, consistent,
certain, clear and accurate
• 5 Important topics to cover in the message
1. Who is providing the message?
2. What should people do?
3. When do people need to act?
4. Where is the emergency taking place?
5. Why do people need to act?

© 2013 Eaton, All Rights Reserved. 35

EASE Evac

© 2012 Eaton Corporation. All rights reserved.

EASE Evac
Voice evacuation design software from AFMG®
• Intuitive tool for designing acoustic mass notification
concepts in a room, hall or building complex
• 3D simulation software calculates the distribution of
direct sound levels as well as total sound levels
• Signal-to-noise ratio (SNR)
• Speech intelligibility (STI, ALCons, CIS)
• Benefits
• Enables designers and installers to meet new intelligibility
code requirements with pre-planning
• Saves time and money by reducing costly post-installation
changes and limits over-design
• Powered by AFMG's proven EASE professional sound
design technology, EASE Evac is scaled specifically for
life safety professionals.
• Wheelock speaker data files are available for
download at www.coopernotification.com

37
© 2013 Eaton, All Rights Reserved. 37
EASE Evac Files
Software Tools

Quick Link on Home Page

Cooper Industries > Cooper Safety > Cooper Notification > Resources > Design Center/Tools > Software Tools

© 2013 Eaton, All Rights Reserved. 38

Intelligibility Solutions Page

Resources Tab
White Paper and Links to Wheelock Exceder LED High Fidelity Speaker line and HPSA Documentation

© 2013 Eaton, All Rights Reserved. 39

Intelligibility References
NFPA 72 2013
• Chapter 3 Definitions
• Chapter 7 Documentation
• Chapter 14 Inspection, Test & Maintenance
• Chapter 18 Notification Appliances
• Chapter 24 Emergency Communication Systems
• Annex D Speech Intelligibility
• Supplements 2 Voice Intelligibility for Emergency Voice/Alarm Communication Systems

Other Documents
• B. Crandall, “The Composition of Speech,” Phys. Rev. 10 ser. 2 (1917): 75.
• Schifiliti, R.P., “Speech Intelligibility”, NEMA Supplement, Fire Protection Engineering, Society of Fire Protection
Engineers, Issue No. 16, Fall 2002.
• Geoffroy, N.A., “Measuring Speech Intelligibility in Voice Alarm Communication Systems”, Master’s Thesis,
Worcester Polytechnic Institute, 5 May 2008.
• Jacob, K.D. & Tyson, T., “Computer-Based Prediction of Speech Intelligibility for Mass Notification Systems”,
SUPDET 2008, Fire Protection Research Foundation, Mar 2008.
• Jacob, K.D., “Understanding Speech Intelligibility and the Fire Alarm Code”, presentation at NFPA WSCE,
Anaheim CA, 14 May 2001.
• Casey Grant, “Intelligibility of Fire Alarm and Emergency Communication Systems” Fire Protection Research
Foundation, Nov. 2008
• Erica D. Kuligowski, NIST Technical Note 1779, “General Guidance on Emergency Communication Strategies for
Buildings,” Fire Research Division, Engineering Laboratory, February 2013.

© 2013 Eaton, All Rights Reserved. 40

Energy Efficiency & Strobe
Technology

© 2012 Eaton Corporation. All rights reserved.

LED Technology

LED Strobes
• Closely controls and manages the light
source and output
• Maximizes human response to a light pulse
• Optimizes light based on parameters of area
and perception of human eye
• Ideal to have light on for longer period of time
vs. intensity of light
• Perception of brightness to the eye depends
on how long the light is on
• Light source life*: 50,000 to 100,000 hours

Xenon Strobes
• Light can’t be controlled
• Flash only
• Light source life*: 1000 to 5000 hours
* Assuming Steady Operation

© 2013 Eaton, All Rights Reserved. 42

Recent Studies on Strobe Lighting
Performance Objectives for Light Sources
Used in Emergency Notification Appliances
• Fire Protection Research Foundation - May 2012
• Background:
• LED and other energy saving technology entering
the marketplace
• Existing requirements are based on relatively short
duration, high peak intensity flashing lights
• NFPA 72 and other standards define a method for
calculating the equivalent or effective intensity of a
flashing light source
• Current calculation method is subjective, does not
produce an exact comparison and only
approximates the perceived brightness for direct
viewing of the light source
• Objective:
• Develop methods and criteria to evaluate
performance of light sources used in emergency
notification appliances for inclusion in NFPA 72
• Recommend specifications for both direct and
indirect signaling as well as identify gaps in the
available information

© 2013 Eaton, All Rights Reserved. 43

FPRF – Light Sources Study
Preliminary Findings
• Present requirements for visual signals may not be sufficient for waking
individuals
• Maximum luminous intensity from a flashing signal may be a more predictive
measure of detection performance under indirect viewing conditions
• When ambient light levels are high, indirect viewing becomes much more
difficult. This might affect placement
• Requirements for white/clear light color in sleeping areas may restrict the
amount of light transmitted by the eyelid compared to amber light
• Ambient light levels can affect the conspicuity of visual signal appliances,
especially for outdoor applications
Research Recommendations
• Awakening Sleeping Persons
• Direct Detection
• Indirect and Peripheral Detection
• Performance-based Alternative (0.0375 fc)

© 2013 Eaton, All Rights Reserved. 44

Recent Studies on Strobe Lighting
Parameters for Indirect Viewing of Visual Signals
Used in Emergency Notification
• Fire Protection Research Foundation (FPRF)
• May 2013 – Presented at NFPA Conference on June 10
• Background:
• Research recommendation from the May 2012 Study
• Effective intensity may not be predictive of visual detection of signal lights
when these are viewed indirectly or in the far‐peripheral field of view
• Observers see the change in illuminance on room surfaces rather than the
flashing light itself when it is not in the central field of view
• Flashing light should increase the illuminance on the opposite wall by at least
7% in order for this increase to be detected reliably
• Objective:
• Conduct a human factors laboratory study to identify whether the 7% increase
in light level can be reliably detected by observers with normal vision.

© 2013 Eaton, All Rights Reserved. 45

FPRF - Contour Plot: Detection

LED

LED’s Higher Pulse

creates greater
opportunity to see
the light.

Xenon

© 2013 Eaton, All Rights Reserved. 46

FPRF - Contour Plot: Ease of Detection
Ratings
+2 Very Easy
+1 Somewhat Easy
0 Neither Easy or
Difficult
-1 Somewhat difficult
-2 Very difficult
-3 Not detected

Human response to
the 100 ms pulse
width was far
superior

© 2013 Eaton, All Rights Reserved. 47

 Questions

Contact Information
Ben Thomas
District Manager – South Central Region
BenjaminPThomas@Eaton.com
Cell: 214-912-4944

© 2013 Eaton, All Rights Reserved. 48