This technical note examines the sensitivity of a finite-duration acceleration command obtained using an undamped zero-vibration (ZV) shaper for a linearized overhead crane. A constant-acceleration maneuver lasting one nominal pendulum period is considered, with the acceleration and duration held fixed when the actual system differs from the design model. Exact terminal responses are derived for a constant cable-length error and for unmodeled equivalent viscous damping. The residual amplitude caused by a small relative cable-length error is proportional to the magnitude of that error, whereas the initial envelope of the damped free response is proportional to the damping ratio. Their leading-order ratio is twice the damping ratio divided by the absolute relative cable-length error. This comparison concerns two separate perturbations under the same prescribed maneuver; it is not a universal ranking of friction and cable-length effects. The analysis distinguishes the mathematical impulse filter from the physical command and clarifies the definition of residual vibration in a damped system. It provides a compact basis for assessing model mismatch while identifying the additional considerations required for nonlinear friction, aerodynamic loading, actuator dynamics, and time-varying cable length.