priiieye

Perception&Psychophysics1983,33(1),75-78TheeffectsofflickeradaptationupontemporalcontrastenhancementFREDERICKL.
KITTERLEandBETTINAL.
BEARDUniversityofToledo,Toledo,OhioTemporalcontrastenhancementreferstothefindingthatintermediate-duration,low-spatial-frequencygratingsareperceivedtohaveagreatercontrastthanlong-durationgratingsofsimilarspatialfrequency.
KitterleandCorwin(1979)suggestedthatthiseffectreflectsprimarilyactivityintransientchannels.
Totestthis,30subjectswereruninoneofthreegroupsconsist-ingofeitheradaptationtoasteadyorflickeringlow-spatial-frequencygratingornoadaptation.
Temporalcontrastenhancementwasfoundforbothsteadyandno-adaptationconditions.
Flickeradaptationabolishedtemporalcontrastenhancement.
Itwassuggestedthatflickeradaptationmaydecreasethecontributionoftransientchannelstoperceivedcontrastandmaycausepro-cessingofthetestgratingstobeswitchedtosustainedchannels.
Theimplicationsofthesere-sultsforunderstandingthecodingofbrightnessarediscussed.
Psychophysicalstudieshaveshownthataflick-eringgratingorlinecanbesettothresholdusingaflickerorpattern-detectioncriterion(Keesey,1972;Kulikowski&Tolhurst,1973;Tolhurst,1973;vanNes,Koenderink,Nas,&Bouman,1967).
Atlowspatialfrequencies,thehumanvisualsystemismoresensi-tivetoflickerthantopattern,whereasathighspatialfrequencies,thereverseisfound.
KulikowskiandTolhurst(1973)haveconcludedthatflickerandpat-ternperceptionaredeterminedbyfunctionallyin-dependentmechanisms.
Theyhavesuggestedthatflickerdetectionismediatedbytransientchannels,whicharemoresensitivetolow-spatial-frequencygratingstemporallymodulatedathighrates,andthatsustainedchannels,whicharemoresensitivetointer-mediate-to-highspatialfrequenciestemporallymod-ulatedatlowrates,mediatepatterndetection.
Burbeck(1981)hasshownthattheseoriginalfind-ingsresultedfromtheuseofthemethodofadjust-ment.
Inthismethod,flickerthresholdsarefoundtobelowerthanpatternthresholdsbecausetheflickertaskrequireslessinformationandthereforeiseasiertodetectthanapattern.
Usingacriterion-freemethodthatrequirestheobservertodistinguishtheteststim-ulusfromastandarddifferinginonlyonedimension(thespatialstructureinthepatternexperimentsortemporalfactorsintheflickerexperiments),Burbeck(1981)hasshownthatpatternsensitivityisequaltoorgreaterthanflickersensitivityforallflickeringgratingsexceptthosethathaveextremelylowspatialfrequencyormoderate-to-hightemporalfrequencies.
Theseresultsdonotsupporttheidentificationofflickerdetectionwithintransientmechanismsandpatterndetectionwithinsustainedmechanisms,Theauthors'mailingaddressis:DepartmentofPsychology,UniversityofToledo,Toledo,Ohio43606.
75whereastheydoindicatethatflickerandpatternde-tectionaremediatedbydifferentphysiologicalmech-anisms.
Inaddition,Green(1981a)hasshownthatthetransientsystemcanmediateperceptionforgratingsupto30cycles/deg.
Thus,recentstudiesquestiontheconclusionofKulikowskiandTolhurst(1973)thatflickerdetectionismediatedsolelybytransientchannelsandpatterndetectionbysustainedchannels.
Nevertheless,theexistenceofthesetwochannelshasbeensupportedbymaskingstudies(Green,1981b;Kitterle,Corwin,&Berta,1979;Mitov,Vassilev,&Manahilov,1981),bydifferencesinadaptabilitytouniformfieldflicker(Green,1981a),andbydiffer-encesincontrastgaincontrol(Burbeck&:Kelly,1981).
Otherstudiesindicatethatthelimitsoftemporalsum-mationareshorterfortransientthanforsustainedchannels(Legge,1978;Nachmias,1967)andthatthelatenciesoftransientchannelsareshorterthanthoseofsustainedchannels(Breitmeyer,1975).
Thesetwomechanismsalsodifferintheirtemporalimpulsere-sponse.
Transientchannelresponsesarebiphasic:theyconsistofanearlyexcitatorycomponentandadelayedinhibitoryresponse(Kelly,1971a,1971b;Ueno,1977;Watson&:Nachmias,1977),whereastheredoesnotappeartobeanydifferenceinthelatencyoftheexcitatoryandinhibitoryresponsesofsustainedchannels(Kitterle,1979;Watson&:Nachmias,1977).
Littleisknownabouttheroleplayedbythesechannelsatsuprathresholdlevelsofillumination.
BrocaandSulzer(1903)haveshownthattheperceivedbrightnessofanintermediate-duration(50-130-msec)flashisgreaterthanthatofalong-durationflashofthesameluminance.
Thisphenomenonhasbeencalledthe"Broca-Sulzereffect"or"temporalbrightnessenhancement.
"Severalmodelsthatemphasizetheroleofneuralresponsetransientsanddelayedinhi-Copyright1983PsychonomicSociety,Inc.
76KITTERLEANDBEARDbitionhavebeenpostulatedtoaccountforthiseffect(e.
g.
,Adrian,1928;Baumgardt&Segal,1947;Boynton,1961).
KitterleandCorwin(1979)testedthepossibilitythattemporalbrightnessenhancementismediatedbytransientchannelsbymeasuringchangesinapparentcontrastforsinusoidalgratingsasafunctionofflashdurationandgratingspatialfrequency.
Itwasfoundthatthefunctionrelatingapparentcontrastanddu-rationvarieswithspatialfrequency.
Apparentcon-trastincreasedwithdurationatarateinverselyrelatedtospatialfrequency,aresultconsistentwiththethresh-oldsummationdataofLegge(1978).
Atlongerdura-tions,however,apparentcontrastbecameindepen-dentofduration,andincreasedataconsiderablyslowerrateforhighspatialfrequencytargets.
Thechangesinapparentcontrastweremorecomplexforthelow-spatial-frequencytargets.
Atlongerdura-tions,apparentcontrastdecreasedandthenbecameindependentofduration.
Thus,theapparentcontrastoflong-durationflasheswaslessthanthatofinter-mediate-durationflashes.
KitterleandCorwin(1979)calledthiseffect"temporalcontrastenhancement"andsuggestedthatthechangesinapparentcontrastfoundwithlow-spatial-frequencytargetsreflectpri-marilyactivityoftransientchannels.
Thepurposeofthisstudywastoprovide,byde-terminingtheeffectsofflickeradaptationupontheapparentcontrastoflow-spatial-frequencytargets,atestofthehypothesisproposedbyKitterleandCorwin(1979).
Recently,Breitmeyer,Levi,andHarwerth(1981)haveshownthatadaptationtoaflickeringuniformfieldof6Hzhasseveraleffectsonvisualperformanceatthresholdandsuprathresholdlevelsofstimulation.
Asaresultofflickeradapta-tion,reactiontimetolow-spatial-frequencygratingsisincreased,theinterstimulusintervalnecessaryforresolvingtwobrieflyflashedsuprathresholdlow-spatial-frequencygratingsincreases,andcontrastsensitivityforthedetectionoflow-spatial-frequencygratingsisreduced.
Breitmeyeret.
al.
(1981)suggestedthattheuniformfieldflickerprimarilyadaptedtran-sientchannels.
Thus,iftransientchannelsmediatetemporalcontrastenhancement,flickeradaptationshoulddiminishorabolishtheeffect.
METHODObserversThirtyundergraduatesparticipatedinthisexperimentforcreditinintroductorypsychology.
Allhadnormalorcorrected-to-normalvisionandwerenaiveaboutthepurposeofthisexperiment.
AppantusThestimuliweresinusoidalgratingsgeneratedbymeansofatwo-channelhapliscopicsystem.
Eachchannelconsistedofanoscilloscope(TektronixSI03/DN,P-31phosphor)uponwhichthegratingsweregeneratedbyconventionalmethods(Campbell&Green,1965).
Bothoscilloscopesweremaskeddowntocircularaperturesthatataviewingdistanceof167emsubtendedavisualangleof3.
4deg,Thescreenswerespatiallyadjacentandseparatedby2deg,Midwaybetweenthedisplayscreenswasadimredfix-ationlight.
Theluminanceofbothscreenswascarefullymatchedto10cd/m',andgratingcontrastwas3401,asmeasuredwithanEG&Gphotometer/radiometer.
ASym-1microprocessor(SynertekCorp.
)wasusedtocontrolthetimingandcollectionofdata.
ProcedureThetaskofthesubjectswastodeterminewhichoftwobrieflyflashedsinusoidalgratingsof.
67cycles/deghadthegreatercon-trast.
Bothofthetargetshadthesamephysicalcontrast,but,onagiventrial,onetargetwaspresentedforTmsecandtheotherfor500msec.
ThistechniqueissimilartothatdescribedbyBowenandPokorny(1978).
Thedatageneratedareplottedtoshowthepercentageoftimethelonger(SOO-msec)flashisjudgedtohaveagreatercontrastasafunctionoftheshorter(Tmsec)flash.
Atveryshortdurations,theSOO-msecflashshouldalwaysbejudgedtobegreaterincontrastthantheshorterflash.
If,however,thereisatemporalcontrastenhancementeffect,therewillbeadurationTforwhichthelongerflashisneverjudgedtohaveagreatercon-trast.
Finally,atsufficientlylongdurations,theT-msecand500-rnsecflashesshouldbejudgednottodifferinapparentcontrast,andthusthejudgmentsshouldasymptotearound50010.
The30subjectsweredividedintothreegroupsofto.
Thereweretwoexperimentalgroups,bothofwhichwerebinocularlyadapted.
Oneofthesegroupsadaptedinitiallyfor2mintoastationarylow-spatial-frequencysinusoidalgratingof.
67cycles/deg.
Theotherexperimentalgroupadaptedtoa6-Hzflickeringgratingof.
67cycles/degforthesameduration.
Thecontrolgroupviewedablankscreenthatwasofthesamemeanluminanceastheadapt-inggratings.
After500rnsec,theadaptationconditionwastermi-natedandthetwotestgratingswerepresented.
Thesubjectpressedoneoftwokeystoindicatewhichofthetwogratingshadthegreaterperceivedcontrast.
Fivehundredmillisecondsafterthisjudgment,theadaptingconditionwasrestoredfor20secandthecyclewasrepeateduntil14replicationshadbeenmadeateachTduration.
RESULTSTheresultsofthisexperimentareshowninFig-ure1.
Thepercentageoftimethatthe500-msecflashisjudgedto.
havemorecontrastisplottedasafunc-tionofthedurationoftheshorterflash.
Thefilledandopencirclesshowtheresultsobtainedfortheun-adaptedandstationary-adaptedconditions,respec-tively.
Theverticalbarsindicatethe95070confidenceinterval.
Thefactthatthetwocurvesarevirtuallyindistinguishableandthatbothdipsignificantlybelow50070withintherangeof60-90msecindicatesthattheBroca-Sulzereffectisnotabolishedbysta-tionaryadaptation.
However,theresultsobtainedwithflickeradaptation(plottedasfilledtriangles)showamonotonicdeclinewithdurationoftheshorterpulse.
Theredoesnotappeartobeanydurationforwhichtheshorterpulseisseentohavemorecontrastthanthe500-msecflash.
Thisindicatesthatflickeradaptationhasabolishedtemporalcontrastenhance-ment.
DISCUSSIONAlthoughtemporalcontrastenhancementisfoundonlywithIow-spatial-frequencygratings(Kitterle&FLICKERADAPTATIONANDTEMPORALCONTRAST77Figure1.
Thepercentoftimethe5OO-msecpulsewasjudgedtohaveagreatercontrastasafunctionofthedurationoftheshorterpulseunderruckeradaptation(filledtriangles),stationaryadapta-tion(opencircles),andnoadaptation(filledcircles).
Eachsymbolrepresentsthemeanof210judgments.
Verticalbarsrepresentthe9SOJoconfidenceinterval.
Corwin,1979;Kitterle&Rysberg,1976),itcannotbeconcludedthattheeffectisnecessarilyduetoactivitysolelywithintransientchannels.
Transientandsus-tainedchannelsoverlapconsiderablyintherangeofspatialfrequenciestowhicheachresponds.
Conse-quently,atsuprathresholdlevels,sustainedchannelactivitymayjointlycontributetothiseffect.
How-ever,Breitmeyeretal.
(1981)haveshownthat,atthresholdandabovethreshold,flickeradaptationmodifiesonlytransientchannelactivity.
Thus,flickeradaptationprovidesameansoftestingtheroleoftransientmechanismsintemporalbrightnessandtemporalcontrastenhancement.
Thepresentpapercomplementstheseearlierpapersandprovidesstrongsupportfortheroleoftransientchannelactivityintemporalcontrastenhancement.
Flickeradaptationappearstohaveselectivelyreducedthecontributionoftransientchannelstotemporalchangesinapparentcontrast.
Ithasbeensuggestedthatflickeradapta-tionoranyotherprocedurethatselectivelyreducestransientactivitymayshiftprocessingtothemoresensitivesustainedmechanisms(Breitmeyeretal.
,1981;Legge,1978).
Thereareseveralfeaturesoftheflicker-adaptedcurve(seeFigure1)thatsuggestthatadaptationmayhavealsoswitchedtheprocessingofthistargettoadifferentmechanism.
First,asmentionedearlier,flickeradaptationabolishestemporalcontrasten-DURATIONOFSHORTERPULSElmsec)hancement.
Thiseffectisnotfoundforstimulithatactivateprimarilythesustainedmechanism.
Second,besidesabolishingtemporalcontrastenhancement,flickeradaptationappearstoshiftthepointatwhichthecurvecrossesthe500'/0linetoaconsiderablylongerduration(i.
e.
,300msec),asopposedto30-40msecfortheothertwoconditions).
Legge(1978)hasshownthatwhentransientchannelactivityismasked,thelimitsoftemporalsummationobtainedwithalow-spatial-frequencygratingincreases.
Legge(1978)sug-geststhatthisresultreflectsachangeinthemecha-nismthatprocessesthegrating(i.
e.
,fromtransienttosustained).
Thus,thepresentdataobtainedunderflickeradaptationmightreflectthetemporalactivityofthesustainedmechanism.
Itisinterestingtonotethat,inarecentlycompletedstudy(Kitterle&Corwin,1983),itwasfoundthatoverarangeoftargetdura-tions,thecurvedescribedbytargetsthatwereturnedongradually(i.
e.
,witharisetimeof20msec)didnotshowevidenceoftemporalcontrastenhance-ment.
Thus,rampingonagratingtargetabolishestemporalcontrastenhancement.
Thisprocedurealsoreducestransientactivityatboththreshold(Breitmeyer&Julesz,1975)andatsuprathresholdlevels(Matsumura,1976).
Thus,itappearsthatexperi-mentalproceduresthatreducethecontributionoftransientactivityalsoreduceorabolishtemporalcontrastenhancement.
Thisfindingratherstronglysuggeststhattransientactivityisaconditionneces-saryforobtainingcontrastenhancement.
Thepresentresultsmayalsohaveimplicationsforunderstandingthemechanismsinvolvedintemporalbrightnessenhancement.
Athighluminancelevels,brightnessenhancementisreducedorabolished(Aiba&Stevens,1964;Magnussen&Bjorklund,1979).
Inviewofthedifferencesincontrastgainofthesetwomechanisms,itisquitepossiblethatattheseluminancelevelsthetransientresponsehassaturated.
Theresultwouldbeashiftintheprocessingofbrightnesstosustainedchannels.
Consistentwiththishypothesisisthefindingthat,atveryhighluminancelevels,thecriticaldurationforthetemporalsummationofbright-nessincreases(Aiba&Stevens,1964).
Itshouldnotbeconcludedthatsustainedchannelsdonotplayaroleinbrightness.
Atlongdurations,apparentcon-trastisgreaterforsquare-wavegratingsthanforsinusoidalgratingsofthesamefundamentalfrequency(Kitterle&Corwin,1979).
Sincesquare-wavegrat-ingscontainhigherharmonics,thesefindingssuggestthatsustainedchannelsmaycodethebrightnessoflong-durationtargets.
Althoughthecandidatecodeforbrightnessisnotknownbecauseitisnotclearhowthesensoryresponsevarieswithtimeforaflashofagivenduration,transientchannelsrespondwithaninitialhighburstofactivityfollowedbylowermaintainedactivity(Oreen,1981b;Mitovetal.
,1981).
Thisisnotthecaseforsustainedchannelactivity,inwhichtheinitialburstisnotfound.
Thus,ifthis200300508010030.
.
FLICKERADAPToSTATIONARYADAPTNOADAPTATION1078KITTERLEANDBEARDparticularneuralfeatureisimportant,anyprocedurethatdiminishesitsmagnitudeshouldalterthebright-nessresponse.
Consequently,itwouldbeofinteresttoexaminetheinfluenceofflickeradaptationuponbrightnessenhancementwithuniformfields.
Thepresentresultssuggestthatsteadyadaptingfieldsshouldhavelittle,ifany,effectontemporalbright-nessenhancement,sincetheydonotadapttransientchannels.
ConsistentwiththisinterpretationaretheresultsofAibaandStevens(1964),whodidnotfindanyeffectofsteadyadaptationonbrightnessen-hancement.
Ifflickeradaptationreducedbrightnessenhancement,thenthistechniquecouldprovidein-sightintothecandidatecodeforbrightness.
Inpar-ticular,thesemeasurescouldtestdirectlythehypoth-esisthattheinitialtransientactivitythataccompaniestheonsetofaflashisaprerequisiteforobtainingbrightnessenhancement.
Thisfindingwouldcontra-dictthosetheoriesthatascribelittleimportancetotheroleofneuralresponsetransientstothisphe-nomenon(Wasserman&Kong,1974).
REFERENCESADRIAN,E.
D.
Thebasisofsensation:Actionofthesenseorgans.
London:Christophers,1928.
AlBA,T.
S.
,&STEVENS,S.
S.
Relationofbrightnesstodurationandluminanceunderliaht-anddark-adaptation.
VisionRe-.
.
rch,1964,4,391-i04.
BAUMGAaDT,E.
,&SEGAL,J.
Facilitationetinhibition,para-metresdelafonctionvisuelle.
AIIIIiePlycholog/qlle,1947,43-44,54-102.
BOWEN,R.
W.
,&POKORNY,J.
Targetedgesharpnessandtem-poralbrightnessenhancement.
VisionResearch,1978,IS,1411-1412.
BOYNTON,R.
M.
Sometemporalfactorsinvision.
InW.
Rosen-blith(Ed.
),&nsorycommunication.
NewYork:Wiley,1961.
BUITIIEYER,B.
Simplereactiontimeasameasureoftransientandsustainedchannels.
VisionResearch,197',15,1411-1412.
BREITMEYER,B.
,&JuLBSZ,B.
Theroleofonandofftransientsindeterminingthepsychophysicalspatialfrequencyresponse.
VisionResearch,197',15,411-41'.
BUITIIEYE"B.
,LEVI,D.
M.
,&HARWERTH,R.
S.
Flickermaskinginspatialvision.
VisionResearch,1981,21,1387-1394.
BmeA,A.
,&SULZEB,P.
Lasensationluminlseenfonctiondutemps.
JourntlldePhyslologleetdePathalogleCHnlrale,1903,6,,,-68.
BUBBECK,C.
A.
Criterion-freepatternandflickerthresholds.
JourntlloftheOpticalSocietyofAmerica,1981,71,1343-1350.
BURBECK,C.
A.
,&KELLY,D.
H.
Contrastpinmeasurementsandthetransient/sustaineddichotomy.
JourntlloftheOpticalSocietyofAmerica,1981,71,1335-1342.
CAMPBELL,F.
W.
,&GUEN,D.
G.
Opticalandretinalfactorsaffectingvisualresolution.
JourntllofPhysiology,196',Ill,576-'93.
GUEN,M.
A.
Psychophysicalrelationshipsamongmechanismssensitivetopattern,motionandflicker.
VilionReMfIrch,1981,21,971-984.
A.
Spatialfrequencyeffectsinmaskingbylight.
Vi-sionResearch,1981,21,861-866.
(b)KEESEY,U.
T.
Flickerandpatterndetection:Acomparisonofthresholds.
JournaloftheOpticalSocietyofAmerica,1972,,62,446-448.
.
KELLY,D.
H.
Theoryofflickerandtransientresponses.
I.
Uni-formfields.
JournaloftheOpticalSocietyofAmerica,1971,61,'37-'46.
H.
Theoryofflickerandtransientresponses.
II.
Counterphasegratings.
JournaloftheOpticalSocietyofAmer-ica,1971,61,632-640.
(b)KITTEBLE,F.
L.
Reactiontimemeasuresofsuprathresholdspatia-temporalinteractions.
SupplementtoInvestigativeOphthal-mologyandVisualScience,1979,18,250.
KITTERLE,F.
L.
,&CORWIN,T.
R.
Enhancementofapparentcontrastinflashedsinusoidalgratings.
VisionResearch,1979,19,33-39:KITTEBLE,F.
L.
,&CORWIN,T.
R.
Theeffectsoftemporalwave-formuponapparentcontrast.
PerceptioncIPlychophyslcs,1983,33,72-74.
KITTEBLE,F.
L.
,&RY8BERG,J.
A.
Theeffectofexposuredura-tionontheapparentcontrastofsinusoidalgratings.
PerceptioncIPsychophysics,1976,I,33'-338.
KITTEBLE,F.
L.
,CORWIN,T.
R.
,&BERTA,J.
Maskingofsinusoidaltargetsbyuniformfieldsofunequalduration.
JourntlloftheOpticalSocietyofAmerica,1979,69,144'.
KULIKOWSKI,J.
J.
,&TOLHURST,D.
J.
Psychophysicalevidenceforsustainedandtransientdetectorsinhumanvision.
JournalofPhysiology,1973,232,149-162.
LEGGE,G.
E.
Sustainedandtransientmechanismsinhumanvi-sion:Temporalandspatialproperties.
VisionResearch,1978,18,69-81.
MAGNUSSEN,S.
,&BJORKLUND,R.
A.
Theperceptionofsupra-thresholdsinusoidalflickermeasuredbylight·anddark-phasematching.
VisionResearch,1979,19,33'-338.
MATSUMURA,M.
Visualmaskingbyluminanceincrementanddecrement:Effectsofrisetimeanddecaytime.
TohokuPsv-chologicaFolia,1976,35,104-114.
MlTOV,D.
,VASSILEV,A.
,&MANAHILOV,V.
Transientandsus-tainedmasking.
Perception&:Psychophysics,1981,30,205-210.
NACHMIAS,J.
Theeffectofexposuredurationonvisualcontrastsensitivitywithsquare-wavegratings.
JournaloftheOpticalSocietyofAmerica,1967,57,421-427.
TOLHURST,D:J.
Separatechannelsfortheanalysisoftheshapeandmovementofamovingvisualstimulus.
JournalofPhys-Iology,1973,231,38'-402.
UENO,T.
Temporalcharacteristicsofthehumanvisualsystemasrevealedbyreactiontimetodoublepulsesoflight.
VisionRe-search,1977,17,'91-'96.
VANNES,F.
L.
,KOENDERINK,J.
J.
,NA8,H.
,&BOUMAN,M.
A.
Spatiotemporalmodulationtransferinthehumaneye.
JourntlloftheOptlealSocietyofAmerica,1967,57,1082-1088.
WAS8EBIIAN,G.
L.
,&KONG,K.
L.
Illusorycorrelationofbright-nessenhancementandtransientsinthenervoussystem.
Science,1974,114,911-913.
WATSON,A.
B.
,&NACHMIA8,J.
Patternsoftemporalinteractioninthedetectionofgratings.
VisionResearch,1977,17,893-902.
(ManuscriptreceivedJune1,1982;revisionacceptedforpublicationOctober20,1982.
)