Abstract
BACKGROUND & PURPOSE A uniform surround can induce changes in the brightness or color of a test region. A surround sinusoidally varied in time in luminance or chromaticity induces perceived temporal variation in a physically constant test field. DeValois et al. (Vision Research 1986) reported strong induction in a test field at inducing-field temporal frequencies below ∼3 Hz but not at higher frequencies. Here we investigated whether nonlinear neural processing occurs prior to the site of neural filtering of temporally-varying induction. If so, simultaneous modulation of two different temporal frequencies above 3 Hz may induce perceptual modulation in the test. For example, a nonlinear process may produce a lower-frequency neural response at the difference (beat) frequency of two simultaneously presented inducing frequencies. METHODS A steady EEW annular test ring (0.5 deg wide) was embedded in a larger circular surround (6 deg diameter). Test and surround were separated by a thin (3 min) dark gap. The surround was modulated (i) sinusoidally at various single temporal frequencies or (ii) with the superposition of two temporal frequencies. The induced temporal modulation in the test ring was matched by adjusting the modulation amplitude of a separate annular “matching ring” varied sinusoidally at 2 Hz. RESULTS As reported previously, temporal induction from sinusoidal surround modulation was severely attenuated above ∼3 Hz. The superposition of two higher temporal frequencies (e.g. 4.17 & 5 Hz), however, induced substantially more temporal variation in the test than either temporal frequency presented alone. CONCLUSION Induced temporal modulation can result from higher temporal inducing frequencies than reported previously. This can be explained by a nonlinearity, which precedes a neural low-pass temporal filter and which produces a neural response at a lower temporal frequency than either inducing stimulus. This lower frequency alone survives a subsequent low-pass filter.
Supported by PHS grant EY-04802.