Amplitude-dependent damping, being nonlinear, has associated with it an acoustic harmonic spectrum. Sufficient data have now accumulated to make desirable an effort to understand the complicated results. We have recently suggested that a sharp dip as a function of amplitude of the third harmonic in iron can be explained as a phase cancellation effect. A simple model, a discontinuous jump in amplitude of a sine wave, indicated that the phase of the emitted acoustic radiation depended on the phase of the jump. Supposing a distribution of breakaway stresses for dislocation segments, one expects that some segments will be emitting with a phase sufficiently different from others to give significant cancellation effects. In this paper, using a more physical representation of the breakaway event, we confirm the results on the tendency of the phases and observe qualitatively that they may be adequate to explain the data for brass as well as iron. Detailed modeling now seems worthwhile.