Abstract

Most frequency-domain nonlinear-optical techniques measure only the magnitude of the Fourier transform of a temporal response and, hence, do not uniquely determine the response. We show that, for the commonly used response, h(t) = A exp(−t/τ_f) + B exp(−t/τ_s), where A ≡ α/τ_f and B ≡ (1 − α)/τ_s, the spectral line shape can always be fitted by two different values of α. A measurement of the optical Kerr transient of carbon disulfide illustrates this ambiguity. We also demonstrate a single-scan method that is free from such ambiguities. It involves adding coherent background with a nonzero quadrature-phase component, obtained simply by proper choice of probe wavelength.

© 1988 Optical Society of America

Narrow nonlinear-optical resonances in CdSSe-doped glass

J. T. Remillard and D. G. Steel
Opt. Lett. 13(1) 30-32 (1988)

Distinguishing molecular-reorientation gratings from thermal gratings by a time-delayed method

Zuhe Yu, Xin Mi, Qian Jiang, Peixian Ye, and Panming Fu
Opt. Lett. 13(2) 117-119 (1988)

5-mm-resolution optical-frequency-domain reflectometry using a coded phase-reversal modulator

David W. Dolfi, Moshe Nazarathy, and Steven A. Newton
Opt. Lett. 13(8) 678-680 (1988)

Temporal reshaping of ultrashort laser pulses after reflection from GaAs at Brewster’s angle

I. H. Campbell and P. M. Fauchet
Opt. Lett. 13(8) 634-636 (1988)

Absolute minimum instability thresholds in a nonlinear ring cavity containing a Kerr medium

W. J. Firth and S. W. Sinclair
Opt. Lett. 13(3) 218-220 (1988)