parameters for indirect viewing of visual signals used in

Parameters for Indirect Viewing of Visual

Signals Used in Emergency Notification

Final Report

Prepared by:

John D. Bullough Nicholas P. Skinner

Yiting Zhu

Lighting Research Center, Rensselaer Polytechnic Institute

© September 2013 Lighting Research Center, Rensselaer Polytechnic Institute

THE FIRE PROTECTION RESEARCH FOUNDATION ONE BATTERYMARCH PARK

QUINCY, MASSACHUSETTS, U.S.A. 02169-7471 E-MAIL: [email protected] WEB: www.nfpa.org/Foundation

iii

FOREWORD

LEDs and other innovative energy saving lighting technologies (e.g. fluorescents) are rapidly entering the marketplace and present themselves for application to emergency notification appliances. The existing requirements for the performance and application of visible notification appliances are based on relatively short duration, high peak intensity flashing lights – strobe lights. NFPA 72, National Fire Alarm and Signaling Code, and referenced listing standards define a method for calculating the equivalent or effective intensity of a flashing light source. The calculation method is subjective and does not produce an exact comparison and is intended only to approximate the perceived brightness for direct viewing of the light source. It has worked because all of the lights approved using the standard have all had relatively similar and short pulse durations. Thus, the peak intensities have been relatively similar. A review of research performed for the Fire Protection Research Foundation by the RPI Lighting Research Center suggested that 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. In particular, 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. Based on previous literature, the previous Foundation study suggested that a flashing light should increase the illuminance on the opposite wall by at least 7% in order for this increase to be detected reliably. For an ambient horizontal illumination level of 100 footcandles (fc) on the work plane in a space such as an office, it was estimated that the vertical illuminance on the wall should increase by at least 2 fc to be reliably detected. This estimate has not been tested empirically. The objective of this project was to conduct a human factors laboratory study to identify whether the 7% increase in light level can be reliably detected by observers with normal vision when viewed indirectly. The results from this study provide technical basis to support the development of methods and criteria to evaluate performance of light sources used in emergency notification appliances for NFPA 72. The Research Foundation expresses gratitude to the report authors John D. Bullough, Nicholas P. Skinner, and Yiting Zhu with the Lighting Research Center at Rensselaer Polytechnic Institute located in Troy, New York. The Research Foundation appreciates the guidance provided by the Project Technical Panelists, the funding provided by the sponsors, and all others that contributed to this research effort. The content, opinions and conclusions contained in this report are solely those of the authors.

Keywords: fire protection building systems, visible notification appliances

v

PROJECT TECHNICAL PANEL

Robert Elliot, FM Approvals

Bruce Fraser, Fraser Fire Protection Services

Warren Olsen, Fire Safety Consultants, Inc.

Lee Richardson, NFPA Staff Liaison

Robert Schifiliti, R.P. Schifiliti and Associates, Inc.

Andrew Trotta, Consumer Product Safety Commission

PROJECT SPONSOR REPRESENTATIVES

Dan Finnegan, Larry Grodsky, and Mark Pavlica, Siemens

Isaac Papier, Honeywell/System Sensor and National Electrical Manufacturers

Association

Rodger Reiswig, SimplexGrinnell

CONTRIBUTING PROJECT SPONSOR REPRESENTATIVES

Jack McNamara, Bosch

Dave Newhouse, Gentex Corporation

vii

TableofContents1.Background..........................................................................................................................................................12.Methods..................................................................................................................................................................3 2.1.ExperimentalLaboratory3 2.2.TestLightSource4 2.3.ExperimentalConditions6 2.4.SubjectsandProcedure73.Results.....................................................................................................................................................................8 3.1.Experiment18 3.2.Experiment29 3.3.Experiment39 3.4.Experiment411 3.5.Experiment511 3.6.Experiment613 3.7.Experiment7134.Discussion...........................................................................................................................................................16 4.1.PartialTemporalSummation16 4.2.IntegrationTimes165.Conclusions........................................................................................................................................................216.Acknowledgments..........................................................................................................................................227.References..........................................................................................................................................................23Appendix:ExperimentalData......................................................................................................................24

1

1.BackgroundVisualsignalingappliancesusedforemergencynotificationhavecommonlyusedxenonstrobelightsthatproduceverybrief(<1ms),high‐intensityflashesoflight,andtheirspecifiedperformanceischaracterizedbytheireffectiveintensity,whichisanestimateoftheluminousintensity(incandelas[cd])ofasteady‐burninglightthathasequivalentvisualeffectivenessastheflashinglight.Evaluationsofvisualsignalsusedinawidevarietyofapplicationshavegenerallyconfirmedtheutilityoftheeffectiveintensityforvisualsignalswhentheyareviewedwithinthecentralportionofthefieldofview(IALA,2008;Bulloughetal.,2013;BulloughandSkinner,2013),regardlessofthespecifictemporalcharacteristicsoftheflashinglight.Theeffectiveintensity(Ie,incd)isdefined(IES,1964)intermsofEquation1: Ie=∫t1t2I(t)dt/(a+t2–t1) (Eq.1)wheret1andt2arethestartandendtimes(inseconds[s])oftheflashoflight,respectively;I(t)istheinstantaneousluminousintensity(incd)oftheflashattimet;andaisaconstantdeterminedempirically(BlondelandRey,1912)tohaveavalueofapproximately0.2s.Thisconstantisrelatedtothetemporalintegrationofthevisualsystemundervisualconditionssimilartothoseunderwhichnavigationalsignallightsarejustatthethresholdfordetectionatnight,whenthevisualsystemisdark‐adapted.Eventhoughmanyvisualsignals,includingthosedeployedinbuildingsforemergencynotificationapplications,arenotviewedunderdarkadaptation,andaredesignedtobeseenatwellabovethresholdlevels,theeffectiveintensityconcepthasheldupforon‐axisandnear‐on‐axisviewingconditions(referredinthisreportas"direct"viewing).InspectionofEquation1suggeststhatthesameeffectiveintensitycanbeachievedwithsignallightshavingverydifferenttemporalintensitycharacteristics.Thedurationoftheflashoflightanditsinstantaneousintensitycanbetradedoff,sothataverybrief,high‐intensityflashcouldhavethesameeffectiveintensityasalonger,lower‐intensityflashoflight.Thishasimplicationsforvisualsignalsusingsolid‐statelightingtechnologysuchaslight‐emittingdiodes(LEDs).LEDsareavailablewithincreasingefficiencyandbrightness,makingthesesourcespracticalforvisualsignalingdevices.Unlikethexenonsourcesinstrobelights,LEDscanbeflashedwithdifferentdurationsandtemporalwaveforms,sothattwoappliancescouldhavethesamecalculatedeffectiveintensitybuthaveverydifferenttemporalflashpatterns(Bulloughetal.,2012a).Areview(Bulloughetal.,2012b)ofpreviousliteraturedescribingtheresultsofstudiesconductedwithxenonstrobelightsources(UL1991)confirmedthatforindirectdetectionwhenthevisualsignalisinornearthefieldofview,aneffectiveintensityof15cdprovidedreliablelevelsofdetection.However,theresultsofthesamestudiesshowedthatflashingincandescentsourcesusedasvisualsignalsrequiredmuchhighereffectiveintensitiestobedetectedreliably.Incandescentflashingsourceshavelongerflashdurationsthatxenonstrobelightsbecauseoftheinherentpropertiesofthefilamentthatmustheatuptoproducelight,andcoolofftostopproducinglight.AmorerecenttestusingLEDsources

2

varyinginintensityanddurationtoachievethesame15cdeffectiveintensity(Savage,2011)reportedthatsourceswithlongerdurationstendedtohavelowerdetectability.Researchstudiesonlargefield,lowfrequencyflickerperception(Kelly,1961)suggestedthatwhenflashingofalargefield(asinindirectviewingofanemergencyvisualsignal)occurredatafrequencynear1‐2Hz,theabsolutemodulationlevelofabout7%producedreliabledetection.Basedonthesefindings,Bulloughetal.(2012b)postulatedthatperhapstheabsoluteorinstantaneousintensityfromasignalwhenviewedindirectlymightbemoremeaningfulthatitseffectiveintensitycharacterizedusingEquation1.ToassisttheFireProtectionResearchFoundation(FPRF)oftheNationalFireProtectionAssociation(NFPA)inidentifyingpredictivemetricsforcharacterizingtheperformanceofvisualsignalshavingtemporalpropertiesdifferentfromxenonstrobes,aseriesofhumanfactorsexperimentswasconductedbytheLightingResearchCenter(LRC);theseexperimentsaredescribedinthepresentreport.

3

2.Methods2.1.ExperimentalLaboratoryThetestlaboratoryusedforthehumanfactorsexperimentwastheSeminarRoomattheLRClaboratoryinTroy,NY.Thisisalargeclassroomspacewithwhitepaintonthreewallsandunpaintedbrickonthefourthwall.Figure1illustratesaschematicdiagramoftheoverallexperimentalset‐up,showingtherelativelocationsoftheexperimentalsubjects,thetestlightsourceandtheoppositewallinthefieldofview(nottoscale).Subjectswereseatedatatablefacingawayfromtheunpaintedbrickwall.Figure2showsthesubjects’viewofthewhitefallfacingthem.

Figure1.Schematicofexperimentallaboratory(positionofsubjectsnottoscale).

Figure2.Viewoffacingwallfromsubjects’position.

Thelightingsystemintheroomisabletobecontrolledthroughaseriesofdimmingswitched.Dependingupontheexperiment,theilluminanceintheseminarroomwasadjustedtoproducethefollowingconditions: Highambientcondition:Averagehorizontalilluminanceontabletopswas500lx,

averageverticalilluminanceonwallfacingsubjectswas200lx

4

Lowambientcondition:Averagehorizontalilluminanceontabletopswas250lx,averageverticalilluminanceonwallfacingsubjectswas100lx

2.2.TestLightSourceFigure3showsthetestlightsource,whichconsistsofanarrayofthreehighpowerwhiteLEDs(Manufacturer:Cree,Partnumber:XREWHT‐L1‐0000)overwhichlenses(Manufacturer:Khatod,Partnumber:KEPL1127)arefitted.Thelensesincludethreepossibletypesofopticaldistributions:40o,25oor6o.(ThebluefilmsvisibleoverthefrontlensaperturesinFigure3areprotectivecoversandareremovedduringtesting.)Thecorrelatedcolortemperature(CCT)rangeofthewhiteLEDswasapproximately3200KinordertomatchtheCCTofthefluorescent(3500K)andhalogen(2800K)lightinginthetestlaboratory.

Figure3.Testlightsource.

Twotypesofoptics(Manufacturer:Khatod)wereusedtogetherwiththeLEDs:onecreatesanarrowbeamangleof6(Partnumber:KEPL112706)whiletheothercreatesawidebeamangleof40(Partnumber:KEPL112740).AsshowninTable1,toincreasetheverticalilluminanceby15lx(anapproximate7%verticalilluminanceincrease)onthefacingwallatheightof6ft,itrequireddrivingcurrenttothewhiteLEDtobe61mAforthe6obeamangleopticand961mAforthe40obeamangleoptic.Figure4illustratesopticalray‐tracingsimulationresultsofdifferentbeampatternsonthewallwithdifferentbeamangleoptics:6and40.TheLEDandtheopticsweremountedatheightof6ft,andtheyareaimingperpendicularlytothewall,whichwas20ftawayfromthelightsource.AsillustratedinFigure3,6opticsilluminatea2ftby2ftarea,whilethe40opticsilluminatealmostthewhole12ftby12ftarea.Therefore,6,and40opticswerechoseninthisexperimenttorepresentnarrowandwidebeamtypes.Thelightsourcewasmountedintheceilingbehindthesubjects’seatingposition.Aplywoodbafflewasusedtoblockaportionofthelightfromthesourcetoavoidilluminatingashadowedportionofthewallfacingthesubjects,showninFigure5.

5

Figure4.Opticalray‐tracingsimulationofverticalilluminanceonthewallwhenusingdifferentbeamangleoptics(6,topand40,bottom)ona12ftby12ftarea

withthelightsourceaimedatthecenter.

Figure5.Photographofthebaffleandlightsourcemountedintheceiling.

6

2.3.ExperimentalConditionsThelightsourceswerecontrolledusingacustomLabview(NationalInstruments)programhatcouldproduceflashesoflightvaryinginintensity(intermsoftheilluminanceincrementonthewallfacingthesubjects’seatingposition)andduration,withflashratesfrom1Hz(oneflashpersecond)to2Hz(oneflashperhalf‐second).Insomeexperiments,subjectswererequestedtolookdirectlyaheadatthewallfacingthem,andinothers,subjectsperformedanumericalverificationtask(NVT)placedonthetableinfrontofthemandwouldhavedetectedtheflashingintheirperipheralvision.Thenumericalverificationtaskconsistedofprintedcolumnsofnearlymatching5‐digitnumbers,with3%ofthedigitsnotmatching.Subjectswereinstructedtoplaceacheckmarknearnon‐matching5‐digitnumbersandtoreportwhethertheysawflashingwhileperformingthistask.Inthefinalexperiment,subjectsunawareofthepurposeoftheexperimentperformedthetaskandthenansweredwhethertheynoticedtheflashingafteranexperimentalconditionwasdisplayed.Table1.Preliminarymeasurementresultsoftheilluminancelevel(“I”represents

drivingcurrent)Height(ft)

VerticalIlluminance(lx)Ambientlightonwall 6optics(I=61mA) 40optics(I=961mA)

8 117 7 206 6 219 15 155 208 4 198

Thesourcewascalibratedandadjustedtoproducethefollowingexperimentalconditionsineachofthesevenexperimentsthatwereconducted: Experiment1:Ambientilluminance500lx;Beamangle40o;Duration1,10,100ms;

Illuminanceincrement1%,2%,4%,8%;Frequency1Hz;Subjectslookingatwall Experiment2:Ambientilluminance500lx;Beamangle40o;Duration10,25,50,100

ms;Illuminanceincrement1%,2%,4%,8%;Frequency2Hz;Subjectslookingatwall Experiment3:Ambientilluminance250lx;Beamangle40o;Duration10,25,50,100

ms;Illuminanceincrement2%,4%,8%,16%;Frequency1Hz;Subjectslookingatwall Experiment4:Ambientilluminance500lx;Beamangle40o;Duration10,25,50,100

ms;Illuminanceincrement1%,2%,4%,8%;Frequency1Hz;SubjectsperformingNVT Experiment5:Ambientilluminance250lx;Beamangle40o;Duration10,25,50,100

ms;Illuminanceincrement2%,4%,8%;16%;Frequency1Hz;SubjectsperformingNVT

Experiment6:Ambientilluminance250lx;Beamangle6o;Duration10,25,50,100ms;Illuminanceincrement2%,4%,8%,16%;Frequency1Hz;SubjectsperformingNVT

Experiment7:Ambientilluminance250lx;Beamangle40o;Duration50ms;Illuminanceincrement4%,16%;Frequency1Hz;SubjectsperformingNVT

7

InExperiments4and5acommerciallyavailableemergencynotificationvisualsignalwasalsoincluded.Thesignalhadanadjustableeffectiveintensitysetting,whichwassetduringbothexperimentstothenominal15cdsetting.Basedonintensityprofiledataprovidedbythemanufacturer,anestimatedvalueof40cdfortheeffectiveintensity(whencalculatedusingEquation1)wasusedforsubsequentanalysis.Duringtheexperimentsinwhichthissignalwasused,itwasfittedwithabaffletoproduceadistributionsimilartothatoftheLEDtestsourcewhenequippedwiththe40olensoptics.2.4.SubjectsandProcedureTensubjectsparticipatedineachexperiment.InExperiments1through6,subjectswereexposedtoeachoftheexperimentalconditionsinarandomorderfor10safterwhichtheywereaskedwhethertheydetectedtheflashinglight.Threenullconditiontrialswithnoflashingpresentwerealsopresentedineachofthesesixexperimentstomeasureresponseswhennosignalwaspresent.InExperiment7,subjectswereshownonecondition(4%or16%illuminanceincrement)only;halfsaweachcondition.InExperiments1through7,subjectsalsoratedtheease/difficultyofdetectionusingthefollowingscale(iftheydidnotdetecttheflashing,avalueof‐3wasassignedastheresponsetothisquestion): +2 Veryeasy +1 Somewhateasy 0 Neithereasynordifficult ‐1 Somewhatdifficult ‐2 VerydifficultInExperiments2through7,subjectsalsoratedtheperceivedurgencyofdetectionusingthefollowingscale(iftheydidnotdetecttheflashing,avalueof‐1wasassignedastheresponsetothisquestion): 3 Veryurgent 2 Somewhaturgent 1 Slightlyurgent 0 Notatallurgent

8

3.ResultsTheresultsofeachexperimentareincludedinthissection,withthedetectionpercentages,ease/difficultyratings,andurgencyratingsforeachconditionplottedasafunctionoftheeffectiveintensityforeachconditioncalculatedusingEquation1.Forthenull‐conditiontrialsinalloftheexperiments,the“falsepositive”detectionratesinalloftheexperimentsrangedfrom0%to3%.Theserateswerelowenoughtofeelconfidentthatsubjectswereonlyveryrarelyrespondingthattheydetectedsomethingwhennoflashingconditionwaspresented,andthatthedetectionpercentagesreportedherearerepresentativeofthelikelihoodthatasignallightwouldbedetected.3.1.Experiment1Figure6ashowsthedetectionpercentagesandease/difficultyratingsforExperiment1.

a.

b. Figure6.a:DetectionpercentagesforExperiment1;b:EaseratingsforExperiment1.ItcanbeseeninFigure6thattheabilitytoseetheflashingsignalsindirectlywasnotwellpredictedbytheireffectiveintensitybasedonEquation1.Norwastheabsoluteilluminanceincrementpredictiveofperformance.Forexample,inFigure6a,the10msflashwitha2%illuminanceincrementyieldedonly20%detection,butthe100msflashwiththesameilluminanceincrementwasdetected90%ofthetime.The1mssignalswerehardlyeverdetected.

9

Basedontheseresults,the1msconditionswereeliminatedfromfutureexperiments,andinordertoprovidemoreresolutionbetween10and100ms,flashdurationsof25and50mswereincluded.Approximately,theflashdurationsforeachdurationincrementdoubledexceptforthestepbetween10and25increment,afactorof2.5increase.3.2.Experiment2Figure7ashowsthedetectionpercentages,Figure7btheease/difficultyratings,andFigure7ctheurgencyratingsforExperiment2.

a. b.

c.

Figure7.a:DetectionpercentagesforExperiment2;b:EaseratingsforExperiment2;c:UrgencyratingsforExperiment2.

AswithExperiment1,neithereffectiveintensitynortheabsoluteilluminanceincrementwaspredictiveofperformancefortheconditionsinExperiment2.Inaddition,althoughtheflashfrequencybetweenExperiments1and2differed(1HzinExperiment1and2HzinExperiment2),thedetectionperformanceandratedease/difficultyfortheconditionscommontobothexperiments(10and100ms,foreachilluminanceincrement)werehighlycorrelatedtoeachother(r2=0.90fordetection,andr2=0.96forease/difficulty)andverysimilarinmagnitude.Basedonthiscorrelation,allsubsequentexperimentsusedaflashrateof1Hz,sincetheredoesnotappeartobeaneffectoffrequencybetween1and2Hz.3.3.Experiment3Figure8ashowsthedetectionpercentages,Figure8btheease/difficultyratings,andFigure8ctheurgencyratingsforExperiment3.

10

a. b.

c.

Figure8.a:DetectionpercentagesforExperiment3;b:EaseratingsforExperiment3;c:UrgencyratingsforExperiment3.

a. b.

c. Figure9.Correlationofdetection(a),ease/difficulty(b),andurgency(c)forcorresponding

conditionsinExperiments2and3.Asexpected,sincetheambientlightlevelinExperiment3waslowerthaninExperiments1and2,theperformanceandsubjectiveratingswerehigher.Comparingtheconditionsfor

11

whichtheflashdurationandtherelativeilluminanceincrement(2%,4%and8%)werecommontobothexperiments,theresultsoftheseexperimentswereverysimilarandstronglycorrelatedwitheachother(Figure9).Thissuggeststhattherelativeilluminanceincrementratherthantheabsoluteincrementismoreimportantfortheperformanceofasignallightwhenviewedindirectly.Nonetheless,thepositivey‐interceptsofthebest‐fittinglinearfunctionsinFigure9suggestthatperformancewasslightlyimprovedunderthehigherambientlevel,forthesamedurationandrelativeilluminanceincrement.3.4.Experiment4Figure10ashowsthedetectionpercentages,Figure10btheease/difficultyratings,andFigure10ctheurgencyratingsforExperiment4.

a.

Experiment 4: 500 lx, 1 Hz (visual task)

0%

20%

40%

60%

80%

100%

1 10 100 1000

Effective Intensity (cd)

Detection (%)

Xenon

10 ms

25 ms

50 ms

100 ms

b.


‐3

‐2

‐1

0

1

2

1 10 100 1000


Mean

Ease Rating

Xenon

10 ms

25 ms

50 ms

100 ms

c.


‐1

0

1

2

3

1 10 100 1000


Mean Urgen

cy Rating

10 ms

25 ms

50 ms

100 ms

Xenon

Figure10.a:DetectionpercentagesforExperiment4;b:EaseratingsforExperiment

4;c:UrgencyratingsforExperiment4.Thedetectionperformanceislowerforthisexperiment,inwhichsubjectsperformedtheNVTduringthestudy,thanforExperiment2,inwhichsubjectswerepermittedtolookdirectlyatthewall.Itcanalsobeseenthatthedetectionandratingsforthexenonsignalwerehigherthanforanyoftheotherconditions.3.5.Experiment5Figure11ashowsthedetectionpercentages,Figure11btheease/difficultyratings,andFigure11ctheurgencyratingsforExperiment5.

12

a.


0%

20%

40%

60%

80%

100%

1 10 100 1000


Detection (%)

10 ms

25 ms

50 ms

100 ms

Xenon

b.


‐3

‐2

‐1

0

1

2

1 10 100 1000


Mean

Ease Rating

10 ms

25 ms

50 ms

100 ms

Xenon

c.


‐1

0

1

2

3

1 10 100 1000


Mean Urgen

cy Rating

10 ms

25 ms

50 ms

100 ms

Xenon

Figure11.a:DetectionpercentagesforExperiment5;b:EaseratingsforExperiment

5;c:UrgencyratingsforExperiment5.

a.

y = 0.68x + 0.11R² = 0.60

0%

10%

20%

30%

40%

50%

60%

70%

80%

0% 20% 40% 60% 80% 100%

500 lx

250 lx

Detection

b.

y = 0.75x ‐ 0.54R² = 0.82

‐4

‐3

‐2

‐1

0

‐3 ‐2 ‐1 0 1 2

500 lx

250 lx

Difficulty

c.

y = 0.79x ‐ 0.01R² = 0.74

‐2

‐1

0

1

‐1 0 1 2 3

500 lx

250 lx

Urgency

Figure12.Correlationofdetection(a),ease/difficulty(b),andurgency(c)forcorresponding

conditionsinExperiments4and5.TheperformanceinExperiment5washigherthanforExperiment4,whichusedahigherambientlightlevel.However,whencomparedforthesamedurationandthesamerelative

13

illuminanceincrementscommontobothexperiments(2%,4%,8%),theresultsweresimilarandreasonablycorrelatedwitheachother(Figure12).Thepositivey‐interceptsinFigure12suggest,asinFigure10,thatperformancewasslightlyimprovedunderthehigherambientlevelforthesamerelativeilluminanceincrement.3.6.Experiment6Figure13ashowsthedetectionpercentages,Figure13btheease/difficultyratings,andFigure13ctheurgencyratingsforExperiment6.

a. b.

c. Figure13.a:DetectionpercentagesforExperiment6;b:EaseratingsforExperiment

6;c:UrgencyratingsforExperiment6.ItcanbeseenthatusingthenarrowbeamlightsourceinExperiment6resultedingenerallylowdetectionperformanceandlowratingsofease/difficultyandurgency.Onlythehighestilluminanceincrement(16%)foradurationof50mswasdetectedatleast50%ofthetime.3.7.Experiment7Figure14ashowsthedetectionpercentages,Figure14btheease/difficultyratings,andFigure14ctheurgencyratingsforExperiment7.

14

a. b.

c. Figure14.a:DetectionpercentagesforExperiment7;b:EaseratingsforExperiment

7;c:UrgencyratingsforExperiment7.

a. b.

c. Figure15.Correlationofdetection(a),ease/difficulty(b),andurgency(c)forcorresponding

conditionsinExperiments5and7.

15

FewerconditionswereusedinExperiment7thaninExperiment5,whichusedcorrespondingconditions.ThisisbecausesubjectsinExperiment7onlyviewedasingleconditioninordertoidentifytheresponsesofunalertedsubjectsunawareofthepurposeoftheexperiment;oncesubjectswereaskedwhethertheydetectedthefirstconditiontheysaw,theywouldbeawareofthenatureoftheexperimentforanysubsequentpresentations.Tocomparetheimplicationsofunalerted,unawaresubjects,thedatainExperiments5and7werecomparedforcorrespondingconditionsinFigure15.Similartothecomparisonsbetweenambientlightlevels,theperformanceandresponsesoftheunalertedsubjectswhowereunawareofthepurposeoftheexperiment,wasgenerallylowerthanforsubjectswhowereawareofthepurposeoftheexperiment,andthey‐interceptvaluesofthebest‐fittingfunctionsinFigure15arepositive.Theexceptiontothispatternwasforthe16%incrementcondition,whichwasdetected100%ofthetimeinbothexperiments,althoughthisconditionwasratedasmoredifficultandlessurgentbytheunalertedsubjects.

16

4.DiscussionAsdescribedintheresultssection,theeffectiveintensity(asdefinedinEquation1)wasnotausefulpredictivemetricfortheperformanceofvisualsignalshavingdifferentdurations.Norwastherelativeilluminanceincrementpredictiveofperformance,asconsideredbyBulloughetal.(2012b).Theeffectiveintensityformulationisbasedontheconceptthattheintensityanddurationofalightsourcecanbetradedoffinordertomaintaindetectionperformance.AsmentionedbytheIALA(2008),theoriginaleffectiveintensityformulationdevelopedbyBlondelandRey(1912)isprimarilyapplicabletothresholddetectionofsignallightsvieweddirectly(on‐axis)fromlongdistances(andseenaspointsourcesoflight)underdarkadaptationconditions.Incomparison,thevisualsignalsinthepresentstudywereviewedunderhighlightlevelscommontoofficeandotherinteriorapplications,andunderindirectviewing,couldfillarelativelylargeportionofthefieldofviewandofteninthevisualperiphery.Undersuchconditions,theuseofEquation1initscurrentformmaynotbewarrantedforspecifyingindirectdetection.Twoconceptswereexploredtodevelopalternativeformulationsfortheeffectivenessofflashinglightswhenviewedindirectlyunderinteriorlightlevels.4.1.PartialTemporalSummationEquation1assumesalinear,proportionaltradeoffbetweeninstantaneousintensityandflashduration.BaumgardtandHillman(1961)found,forrelativelylarge,7.5o‐diameterluminousstimulipresented20ooffaxisunderdarkconditions,thatforshortdurationstheintensityanddurationcouldbetradedoffproportionallysothatthethresholdoccurredwhentheproductoftheintensityanddurationwasaconstantvalue.Forlongerflashdurations,theproductofthesquareorcuberootoftheintensityanddurationwasconstantforthesamelevelofdetection.InEquation1,thetermI(t)dtwasraisedtoanexponenthavingdifferentvalueslowerorgreaterthan1wasusedtoestimateapossibleindirecteffectivenessquantity.Noneofthevaluesresultedinrelationshipsbetweentheindirecteffectivenessquantityanddetectionperformanceorratedease/difficultyorurgency,thatappearedmuchdifferentfromthegraphsfortheexperimentaldatashowninChapter3(Results)ofthisreport.Thissuggeststhattheconceptofpartialtemporalsummationwasnotapplicabletothedatapresentedhere.4.2.IntegrationTimesAnotheraspectofEquation1thatpertainstothespecificexperimentalconditionsusedbyBlondelandRey(1912)intheirstudiesunderlyingtheeffectiveintensityformulationisthevalueoftheconstantaintheequation.Avalueforaof0.2swasempiricallydeterminedby

17

BlondelandRey(1912).Theconstantaisthoughttoberelatedtothetemporalintegrationtimeofthevisualsystemduringwhichintensityanddurationcouldbetradedoffforequivalentperformanceatdetectionthresholdunderdarkconditionswhentheprimaryphotoreceptorsusedforvisionarerods,whichhaverelativelylongintegrationtimesof0.1‐0.2s.FortheconditionsusedbyBlondelandRey(1912),theconditionsincludedpoint‐sourcesizesignals,viewedunderverydarkconditionsusingon‐axisvision,whenthesignalswerejustbarelyabletobedetected.Othervaluesforahavebeenfoundinvariousstudies,buttherehasbeenrelativelylittlesystematicinvestigationbetweentheoptimalvalueforaandspecificviewingconditionsdifferingfromthoseusedbyBlondelandRey(1912).OneexemplarystudywasconductedbySchmidt‐Clausen(1971),whoinvestigatedtheroleofbackgroundluminance,sizeofthesignallight,andeccentricityinthefieldofviewontheoptimalvalueofainEquation1.Schmidt‐Clausen(1971)foundthatlowervaluesofawerefoundforhigherbackgroundluminances,largersignallightsizes,andgreatereccentricities.ThisisconsistentwithdatafromBattersbyandSchuckman(1970)whoreportedthattemporalintegrationtimesforconephotoreceptors,whicharetheprimaryreceptorsthevisualsystemusesunderdaytimelightlevels,wereontheorderof0.01s.ModificationofthevalueofainthedenominatorofEquation1found,consistentwiththeshorterintegrationtimesofthevisualsystemunderhighlightlevels(BattersbyandSchuckman,1970),andwiththeexperimentaldatafromSchmidt‐Clausen(1971).ThedatafromExperiments1through6werecomparedwithdifferentvaluesofaanditwasfoundthatavalueforaof0.01sresultedinmanyofthecurvesfordifferentsignaldurationsbeingsuperimposedovereachother.Figure16showsthedetectiondataforExperiments1through6(dataforExperiments1and2werecombinedsincetherewasnodifferencebetweenthe1Hzand2Hzdataintheseexperiments),Figure17showstheease/difficultydata,andFigure18showstheurgencydatawhenEquation1ismodifiedusingavalueof0.01sforainthedenominator.Inspectionofthesefiguresshowsthatthecurvesarelocatedclosertooneanotherthanthecurvesinthepreviouschapterofthisreport(Results)andinmanycasesyieldveryconsistentpredictionsfortheflashesoflightwithdifferentdurations.Inaddition,thecurvesseemtoalsobeconsistentwiththedataforthexenonsourceinExperiments4and5inFigures16through18.ThisalsosuggeststhatanindirecteffectivenessquantitybasedonEquation1,butusingavalueof0.01sfora,maybeausefulpredictivemetricforperformance.

18

a.

Detection: Experiments 1‐2

0%

20%

40%

60%

80%

100%

1 10 100 1000a=0.01 s

Detection

1 ms

10 ms

25 ms

50 ms

100 ms

b.

Detection: Experiment 3

0%

20%

40%

60%

80%

100%

1 10 100 1000a=0.01 s

Detection

10 ms

25 ms

50 ms

100 ms

c.


0%

20%

40%

60%

80%

100%

1 10 100 1000

a=0.01 s

Detection

10 ms

25 ms

50 ms

100 ms

Xenonstrobe

d.


0%

20%

40%

60%

80%

100%

1 10 100 1000

a=0.01 s

Detection

10 ms

25 ms

50 ms

100 ms

Xenonstrobe

e.


0%

20%

40%

60%

80%

100%

1 10 100 1000

a=0.01 s

Detection

10 ms

25 ms

50 ms

100 ms

Figure16.ExperimentaldetectiondataforExperiments1‐6plottedasafunctionof

quantitiesusingEquation1withavalueforaof0.01s.

19

a.

Difficulty: Experiments 1‐2

‐3

‐2

‐1

0

1

2

1 10 100 1000

a=0.01 s

Difficulty

1 ms

10 ms

25 ms

50 ms

100 ms

b.

Difficulty: Experiment 3

‐3

‐2

‐1

0

1

2

1 10 100 1000

a=0.01 s

Difficulty

10 ms

25 ms

50 ms

100 ms

c.


‐3

‐2

‐1

0

1

2

1 10 100 1000a=0.01 s

Difficulty

10 ms

25 ms

50 ms

100 ms

Xenonstrobe

d.


‐3

‐2

‐1

0

1

2

1 10 100 1000

a=0.01 s

Difficulty

10 ms

25 ms

50 ms

100 ms

Xenonstrobe

e.


‐3

‐2

‐1

0

1

2

1 10 100 1000a=0.01 s

Difficulty

10 ms

25 ms

50 ms

100 ms

Figure17.Experimentalease/difficultydataforExperiments1‐6plottedasa

functionofquantitiesusingEquation1withavalueforaof0.01s.

20

a.

Urgency: Experiments 1‐2

‐1

0

1

2

3

1 10 100 1000

a=0.01 s

Urgency

10 ms

25 ms

50 ms

100 ms

b.

Urgency: Experiment 3

‐1

0

1

2

3

1 10 100 1000

a=0.01 s

Urgency

10 ms

25 ms

50 ms

100 ms

c.


‐1

0

1

2

3

1 10 100 1000

a=0.01 s

Urgency

10 ms

25 ms

50 ms

100 ms

Xenonstrobe

d.


‐1

0

1

2

3

1 10 100 1000a=0.01 s

Urgency

10 ms

25 ms

50 ms

100msXenonstrobe

e.


‐1

0

1

2

3

1 10 100 1000

a=0.01 s

Urgency

10 ms

25 ms

50 ms

100 ms

Figure18.ExperimentalurgencydataforExperiments1‐6plottedasafunctionof

quantitiesusingEquation1withavalueforaof0.01s.

21

5.ConclusionsThepresentexperimentalresultsandanalysesconductedusingvariationsontheformulationforeffectiveintensitysuggestthattheconventionaleffectiveintensityformulationinEquation1,usingaconstantvalueof0.2sinthedenominator,isnotasuitablemetricforpredictingtheperformanceofasignallightviewedindirectly.Noristherelativeinstantaneousilluminanceincreaseausefulmetric.Itappearsthatausefulpredictivemetric(herecalledanindirecteffectivenessquantity)canbedevelopedfromEquation1,butusingasmallervaluefortheconstanta.Settinga=0.01sallowsthedataforflashesoflighthavingdifferentdurationstobesuperimposedovereachother.Asaroughapproximation,andusingtheperformancewhensubjectswereawareofthenatureoftheexperimentbutwereperformingtheNVTandnotlookingatthefacingwall,theindirecteffectivenessquantityneedstobeapproximately500cdwhentheambientilluminanceis500lx,and250cdwhentheambientilluminanceis250lx,inordertoachieveadetectionpercentageof75%‐80%.[Inordertoachieveadetectionpercentageof90%,acriterionusedbyUL(1991),theindirecteffectivenessquantityshouldbe750cdforanambientilluminanceof500lx,and375cdforanambientilluminanceof250lx.]Thesevaluesfortheindirecteffectivenessquantities(foradetectionlevelof~75%‐80%)alsoresultinameanease/difficultyratingnearzero(theborderlinebetweeneasyanddifficulttodetect)andameanurgencyratingofaboutone(slightlyurgent).Thusitseemsthatachievingthesevaluesorhigherwithasignallightthatisviewedindirectlywouldbegintoensurethatthesignalswereeasytodetectandinterpretedasurgent.Ofcourse,responsesfortheunalerted,naïvesubjectsinthefinalexperimentwerelowerthanfortheprevioussixexperimentsandthismaybeaconsiderationforsettingperformancespecifications.Inaddition,theresponsedataforthenarrowbeamconditionswerequitepoorintermsofdetectionpercentagesandrateddifficultyandurgency.Basedonthepresentdataanemergencynotificationsignalintendedtobeviewedindirectlyshouldilluminatearelativelylargerareaoftheroomsurfaces(e.g.,producingabeamangleofatleast40o)inordertobereliablydetected.ThepresentstudyusedanLEDlightsourcewithaCCTof3200K,matchingtheambientilluminationinthetestlaboratory.ThexenonsourcetestedhasaCCTcloserto5000K.AlthoughtheresultsfromthepresentstudydonotsuggestthedifferenceinCCTbetweenthesesourcesmadealargedifferenceindetection,muchlargerchromaticitydifferencessuchastheuseofcoloredillumination,mightbeeasiertodetectthanthenominally"white"lightsourcesemployedinthepresentexperiments.

22

6.AcknowledgmentsThisstudywassponsoredbytheFireProtectionResearchFoundationoftheNationalFireProtectionAssociation,andperformedunderthedirectionofprojectmanagerAmandaKimball.HelpfulreviewandfeedbackwasprovidedbyRobertElliot,DanFinnegan,BruceFraser,LarryGrodsky,JackMcNamara,DaveNewhouse,WarrenOlsen,IsaacPapier,MarkPavlica,RodgerReiswig,LeeRichardson,RobertSchifiliti,andAndrewTrotta.TheauthorsalsoacknowledgethetechnicalcontributionstothisreportofDr.N.NarendranfromtheLightingResearchCenteratRensselaerPolytechnicInstitute.

23

7.References

BattersbyWS,SchuckmanH.1970.Thetimecourseoftemporalsummation.VisionResearch10(3):263‐73.BaumgardtE,HillmanB.1961.Durationandsizeasdeterminantsofperipheralretinalresponse.JournaloftheOpticalSocietyofAmerica51:340‐344.BlondelA,ReyJ.1912.Theperceptionoflightsofshortdurationattheirrangelimits.TransactionsoftheIlluminatingEngineeringSociety7(8):625‐662.BulloughJD,ZhuY,NarendranN.2012a.Characteristicsoflight‐emittingdiodesources:Relevanceforvisualsignaldetection.Suppression,DetectionandSignalingResearchandApplications:ATechnicalWorkingConference,Phoenix,AZ,March5‐8.Quincy,MA:FireProtectionResearchFoundation,NationalFireProtectionAssociation.BulloughJD,ZhuY.2012b.PerformanceObjectivesofLightSourcesUsedinEmergencyNotificationAppliances[reporttotheFireProtectionResearchFoundation].Quincy,MA:FireProtectionResearchFoundation.BulloughJD,SkinnerNP,TarantaRT.2013.Characterisingtheeffectiveintensityofmultiple‐pulseflashingsignallights.LightingResearchandTechnology45(3):377‐390.BulloughJD,SkinnerNP.2013.Conspicuityofflashesoflight:Interactionsbetweenintensityandduration.JournalofModernOptics[inpress].InternationalAssociationofLighthouseAuthorities.2008.MarineSignalLights,Part4:DeterminationandCalculationofEffectiveIntensity,IALARecommendationE‐200‐4.St.Germain‐en‐Laye,France:InternationalAssociationofLighthouseAuthorities.IlluminatingEngineeringSociety.1964.IESGuideforCalculatingtheEffectiveIntensityofFlashingSignalLights.NewYork,NY:IlluminatingEngineeringSociety.KellyDH.1961.Visualresponsestotime‐dependentstimuli:I.Amplitudesensitivitymeasurements.JournaloftheOpticalSocietyofAmerica51(4):422‐429.SavageK.2011.Flashpulsewidtheffectivenessinnotificationappliances.SUPDET2011Conference.Quincy,MA:FireProtectionResearchFoundation.Schmidt‐ClausenHJ.1971.Theinfluenceoftheangularsize,adaptationluminance,pulseshapeandlightcolurontheBlondel‐Reyconstanta.Theperceptionandapplicationofflashinglights95‐111.AdamHilger,London.

UnderwritersLaboratories.1991.Emergencysignalingdevicesforusebythehearingimpaired,UL1971.Northbrook,IL:UnderwritersLaboratories.

24

Appendix:ExperimentalData

Datafromallsevensetsofexperimentsaresummarizedinthissppendix.Foreachexperiment,themeandetectionpercentagesandthemeansandstandarderrorsofthemean(S.E.M.)forthesubjectiveratings(easeandurgency)arelisted.Experiment1:Ambientilluminance500lx;Beamangle40o;Duration1,10,100ms;Illuminanceincrement1%,2%,4%,8%;Frequency1Hz;Subjectslookingatwall

Mean detection Duration (ms)

1 10 100

Illuminance

increm

ent 1% 10% 0% 30%

2% 0% 20% 90%

4% 0% 90% 100%

8% 0% 100% 100%

Mean Ease Rating Duration (ms)

1 10 100

Illuminance

increm

ent 1% ‐2.9 ‐3.0 ‐2.7

2% ‐3.0 ‐2.7 ‐1.0

4% ‐3.0 ‐0.7 0.4

8% ‐3.0 0.5 1.2

S.E.M. of Ease

Rating

Duration (ms)

1 10 100

Illuminance

increm

ent 1% 0.10 0.00 0.15

2% 0.00 0.21 0.39

4% 0.00 0.54 0.45

8% 0.00 0.34 0.39

Experiment2:Ambientilluminance500lx;Beamangle40o;Duration10,25,50,100ms;Illuminanceincrement1%,2%,4%,8%;Frequency2Hz;Subjectslookingatwall


10 25 50 100

Illuminance

increm

ent 1% 0% 10% 20% 10%

2% 40% 70% 90% 70%

4% 100% 100% 100% 100%

8% 100% 100% 100% 100%

25

Mean Ease Rating

Duration (ms)

10 25 50 100

Illuminance

increm

ent 1% ‐3.0 ‐2.9 ‐2.5 ‐2.9

2% ‐2.4 ‐1.4 ‐0.4 ‐1.7

4% ‐0.4 0.6 0.7 0.7

8% 0.8 1.4 1.5 1.5

S.E.M. of Ease

Rating

Duration (ms)

10 25 50 100

Illuminance

increm

ent 1% 0.00 0.10 0.40 0.10

2% 0.27 0.45 0.52 0.33

4% 0.48 0.31 0.42 0.37

8% 0.47 0.31 0.31 0.31

Mean Urgency

Rating

Duration (ms)

10 25 50 100

Illuminance

increm

ent 1% ‐1.0 ‐0.9 ‐0.6 ‐0.9

2% ‐0.5 0.1 0.4 0.0

4% 0.9 1.2 1.2 0.9

8% 1.4 1.8 1.7 1.6

S.E.M. of

Urgency Rating

Duration (ms)

10 25 50 100

Illuminance

increm

ent 1% 0.00 0.10 0.31 0.10

2% 0.22 0.28 0.27 0.26

4% 0.28 0.29 0.25 0.23

8% 0.37 0.33 0.33 0.27

Experiment3:Ambientilluminance250lx;Beamangle40o;Duration10,25,50,100ms;Illuminanceincrement2%,4%,8%,16%;Frequency1Hz;Subjectslookingatwall


10 25 50 100

Illuminance

increm

ent 2% 20% 90% 70% 70%

4% 80% 100% 100% 90%

8% 100% 100% 100% 100%

16% 100% 100% 100% 100%

26

Mean Ease Rating

Duration (ms)

10 25 50 100

Illuminance

increm

ent 2% ‐2.8 ‐1.5 ‐1.6 ‐1.6

4% ‐0.9 ‐0.2 ‐0.1 ‐0.5

8% 0.0 0.8 0.7 0.8

16% 1.1 1.6 1.7 1.7

S.E.M. of Ease

Rating

Duration (ms)

10 25 50 100

Illuminance

increm

ent 2% 0.13 0.37 0.48 0.48

4% 0.48 0.36 0.38 0.50

8% 0.45 0.42 0.33 0.36

16% 0.38 0.31 0.21 0.21

Mean Urgency

Rating

Duration (ms)

10 25 50 100

Illuminance

increm

ent 2% ‐0.8 0.1 ‐0.1 0.0

4% 0.2 0.6 0.8 0.7

8% 0.9 1.6 1.3 1.1

16% 1.3 2.2 2.3 2.2

S.E.M. of

Urgency Rating

Duration (ms)

10 25 50 100

Illuminance

increm

ent 2% 0.13 0.18 0.23 0.30

4% 0.29 0.22 0.25 0.30

8% 0.31 0.31 0.26 0.28

16% 0.26 0.36 0.33 0.25



10 25 50 100

Illuminance

increm

ent 1% 0% 0% 0% 0%

2% 0% 20% 0% 10%

4% 10% 20% 40% 40%

8% 50% 70% 60% 70%

27

Mean Ease Rating

Duration (ms)

10 25 50 100

Illuminance

increm

ent 1% ‐3.0 ‐3.0 ‐3.0 ‐3.0

2% ‐3.0 ‐2.7 ‐3.0 ‐2.9

4% ‐2.5 ‐2.8 ‐1.9 ‐2.0

8% ‐1.7 ‐0.9 ‐0.9 ‐0.9

S.E.M. of Ease

Rating

Duration (ms)

10 25 50 100

Illuminance

increm

ent 1% 0.00 0.00 0.00 0.00

2% 0.00 0.21 0.00 0.10

4% 0.40 0.13 0.53 0.47

8% 0.47 0.60 0.64 0.55

Mean Urgency

Rating

Duration (ms)

10 25 50 100

Illuminance

increm

ent 1% ‐1.0 ‐1.0 ‐1.0 ‐1.0

2% ‐1.0 ‐0.8 ‐1.0 ‐0.9

4% ‐0.7 ‐0.7 ‐0.2 ‐0.4

8% 0.0 0.4 0.5 0.6

S.E.M. of

Urgency Rating

Duration (ms)

10 25 50 100

Illuminance

increm

ent 1% 0.00 0.00 0.00 0.00

2% 0.00 0.13 0.00 0.10

4% 0.21 0.21 0.33 0.27

8% 0.37 0.37 0.45 0.43

Experiment5:Ambientilluminance250lx;Beamangle40o;Duration10,25,50,100ms;Illuminanceincrement2%,4%,8%;16%;Frequency1Hz;SubjectsperformingNVT


10 25 50 100

Illuminance

increm

ent 2% 0% 0% 20% 0%

4% 30% 10% 60% 30%

8% 20% 60% 90% 60%

16% 80% 100% 100% 90%

28

Mean Ease Rating

Duration (ms)

10 25 50 100

Illuminance

increm

ent 2% ‐3.0 ‐3.0 ‐2.6 ‐3.0

4% ‐2.6 ‐2.8 ‐1.5 ‐2.3

8% ‐2.3 ‐1.1 0.2 ‐1.0

16% ‐0.8 0.9 1.4 0.8

S.E.M. of Ease

Rating

Duration (ms)

10 25 50 100

Illuminance

increm

ent 2% 0.00 0.00 0.27 0.00

4% 0.22 0.20 0.45 0.37

8% 0.47 0.62 0.47 0.60

16% 0.47 0.35 0.16 0.51

Mean Urgency

Rating

Duration (ms)

10 25 50 100

Illuminance

increm

ent 2% ‐1.0 ‐1.0 ‐0.7 ‐1.0

4% ‐0.6 ‐0.8 ‐0.1 ‐0.7

8% ‐0.7 0.1 1.1 0.3

16% 0.4 1.6 1.7 1.5

S.E.M. of

Urgency Rating

Duration (ms)

10 25 50 100

Illuminance

increm

ent 2% 0.00 0.00 0.21 0.00

4% 0.22 0.20 0.28 0.15

8% 0.21 0.35 0.31 0.45

16% 0.27 0.22 0.15 0.37



10 25 50 100

Illuminance

increm

ent 2% 0% 0% 20% 0%

4% 0% 10% 0% 10%

8% 20% 0% 20% 10%

16% 30% 20% 50% 20%

29

Mean Ease Rating

Duration (ms)

10 25 50 100

Illuminance

increm

ent 2% ‐3.0 ‐3.0 ‐2.8 ‐3.0

4% ‐3.0 ‐2.9 ‐3.0 ‐2.9

8% ‐2.6 ‐3.0 ‐2.8 ‐2.8

16% ‐2.6 ‐2.4 ‐1.4 ‐2.5

S.E.M. of Ease

Rating

Duration (ms)

10 25 50 100

Illuminance

increm

ent 2% 0.00 0.00 0.13 0.00

4% 0.00 0.10 0.00 0.10

8% 0.27 0.00 0.13 0.20

16% 0.22 0.43 0.58 0.40

Mean Urgency

Rating

Duration (ms)

10 25 50 100

Illuminance

increm

ent 2% ‐1.0 ‐1.0 ‐0.8 ‐1.0

4% ‐1.0 ‐0.9 ‐1.0 ‐0.8

8% ‐0.7 ‐1.0 ‐0.8 ‐0.9

16% ‐0.6 ‐0.6 0.2 ‐0.7

S.E.M. of

Urgency Rating

Duration (ms)

10 25 50 100

Illuminance

increm

ent 2% 0.00 0.00 0.13 0.00

4% 0.00 0.10 0.00 0.20

8% 0.21 0.00 0.13 0.10

16% 0.22 0.31 0.42 0.21

Experiment7:Ambientilluminance250lx;Beamangle40o;Duration50ms;Illuminanceincrement4%,16%;Frequency1Hz;SubjectsperformingNVT

Duration=50ms Mean

detection

Illuminance

increm

ent

4% 0%

16% 100%

30

Duration=50 ms Mean Ease Rating

S.E.M. of Ease Rating

Illuminance

increm

ent

4% ‐3.0 0.0

16% 1.0 0.5

Duration=50 ms Mean Urgency Rating

S.E.M. of Urgency Rating

Illuminance

increm

ent

4% ‐1.0 0.0

16% 1.0 0.4

parameters for indirect viewing of visual signals used in

Documents