Cochlear nonlinearity between 500 and 8000 Hz in listenerswith normal hearingEnrique A. Lopez-Povedaa)Centro Regional de Investigaciones Biome´dicas, Facultad de Medicina, Universidadde Castilla-La Mancha, Campus Universitario, 02071 Albacete, SpainChristopher J. Plack and Ray MeddisDepartment of Psychology, University of Essex, Wivenhoe Park, Colchester CO4 3SQ, United Kingdom共Received 20 April 2002; accepted for publication 11 November 2002兲Cochlear nonlinearity was estimated over a wide range of center frequencies and levels in listenerswith normal hearing, using a forward-masking method. For a fixed low-level probe, the masker levelrequired to mask the probe was measured as a function of the masker-probe interval, to produce atemporal masking curve 共TMC兲. TMCs were measured for probe frequencies of 500, 1000, 2000,4000, and 8000 Hz, and for masker frequencies 0.5, 0.7, 0.9, 1.0 共on frequency兲, 1.1, and 1.6 timesthe probe frequency. Across the range of probe frequencies, the TMCs for on-frequency maskersshowed two or three segments with clearly distinct slopes. If it is assumed that the rate of decay ofthe internal effect of the masker is constant across level and frequency, the variations in the slopesof the TMCs can be attributed to variations in cochlear compression. Compression-ratio estimatesfor on-frequency maskers were between 3:1 and 5:1 across the range of probe frequencies.Compression did not decrease at low frequencies. The slopes of the TMCs for the lowest frequencyprobe 共500 Hz兲 did not change with masker frequency. This suggests that compression extends overa wide range of stimulus frequencies relative to characteristic frequency in the apical region of thecochlea. © 2003 Acoustical Society of America. 关DOI: 10.1121/1.1534838兴PACS numbers: 43.66.Dc, 43.66.Mk 关MRL兴I. INTRODUCTIONThe mammalian cochlear response is nonlinear inhealthy animals 共Rhode, 1971; Sellick et al., 1982; Robleset al., 1986兲. An increase in the magnitude of stimulationdoes not always produce a proportional increase in the ve-locity or displacement of basilar membrane 共BM兲 vibration.It is generally accepted that for high characteristicfrequencies1共CFs兲 the response is nonlinear for frequenciesclose to CF, but linear for frequencies an octave below CF共Robles et al., 1986兲.Using physiological techniques, cochlear responses havebeen measured in animals in terms of BM input/output 共IO兲functions for a wide range of CFs, stimulation frequencies,and levels 共e.g., Sellick et al., 1982; Robles et al., 1986;Rhode and Cooper, 1996; Recio and Rhode, 2000; Rhodeand Recio, 2000兲. The aim of the present study was to usepsychophysical techniques to estimate the characteristics ofthe human cochlear response over a similar range of param-eters.The nonlinear properties of the human cochlear responsecan be inferred from threshold measurements of maskedprobe tones 共for a review see Moore, 1997兲. A number ofstudies 共e.g., Oxenham and Plack, 1997; Rosen et al., 1998;Baker et al., 1998; Glasberg et al., 1999; Hicks and Bacon,1999; Plack and Oxenham, 2000; Wojtczak et al., 2001; Nel-son et al., 2001; Moore et al., 2002兲 have characterized co-chlear nonlinearity in normal-hearing listeners using this ap-proach. In the present study, a revised version of the methodof Nelson et al. 共2001兲 was used.The method developed by Nelson et al. consists of mea-suring the level of a pure-tone forward masker required tojust mask a pure-tone probe as a function of the masker-probe time interval. The level of the probe is fixed just aboveabsolute threshold. It is thought that the masker level atthreshold depends on two variables. First, it depends on themasker-probe interval: the amount of masking decreases asthe masker-probe interval increases 共Zwislocki et al., 1959;Duifhuis, 1973; Moore and Glasberg, 1983; Nelson andFreyman, 1987兲. Second, it depends on the relative excitationproduced by the masker and the probe at the place on theBM tuned close to the probe frequency 共Oxenham andMoore, 1995; Oxenham et al., 1997; Oxenham and Plack,1997; Nelson et al., 2001兲. Because the probe level is fixedat all times, the method is assumed to measure the maskerlevel 共input兲 required to generate a fixed level of excitationafter decaying during the masker-probe interval. This is thereason that the resulting functions are referred to as iso-response temporal masking curves 共TMCs兲.Obviously, higher masker levels are required as themasker-probe interval increases. However, the slope of theTMC depends on the masker frequency. It has been argued共Nelson et al., 2001兲 that this is because on-frequencymaskers are subject to cochlear compression while others areprocessed more linearly. Therefore, the slope of the TMCsreflects the amount of compression for a given masker. Nel-son et al. showed this behavior for a probe frequency of 1kHz and a wide range of masker frequencies. By assumingthat the internal effect of the masker decays at the same ratea兲Electronic mail: [email protected]. Acoust. Soc. Am. 113 (2), February 2003 0001-4966/2003/113(2)/951/10/$19.00 © 2003 Acoustical Society of AmericaDownloaded 10 Feb 2011 to 65.112.5.1. Redistribution subject to ASA license or copyright; see http://asadl.org/journals/doc/ASALIB-home/info/terms.jspregardless of masker frequency, and that maskers well belowthe probe frequency yield a linear cochlear response, theyderived human cochlear IO curves at CF⬃1000 Hz by plot-ting the masker levels for the low-frequency masker 共a linearreference兲 as a function of the masker levels for other maskerfrequencies.This approach has some advantages over previous meth-ods 共e.g., Oxenham and Plack, 1997; Rosen et al., 1998;Baker et al., 1998; Plack and Oxenham, 2000兲. Fixing theprobe level almost guarantees that the region of the cochleaunder study is the same for different masker 共input兲 levels.Furthermore, fixing the probe level just above threshold en-sures that the CF of the cochlear region under study is closeto the probe frequency. In other words, the effects of ‘‘off-frequency listening’’ are minimized.2In the present study, TMCs were measured for probefrequencies from 500 to 8000 Hz, and for a range of maskerfrequencies at each probe frequency. It will be argued that forlow probe frequencies, cochlear responses are compressedfor maskers well below the probe frequency. This under-mines the assumptions of