This paper presents a comparative investigation of two versatile error-mitigation techniques, applicable to general antenna near-field measurement scenarios with echo signals of unknown origin. Both techniques are based on spatial filtering of the measured field, taking advantage of a priori knowledge of the antenna's size. The first approach takes advantage of the spatial-filtering properties of the spherical-wave expansion of the measured field. The second approach is based on the reconstruction of equivalent currents, and implements the spatial filtering as a direct consequence of the selected size and shape of the reconstruction surface. The investigation was performed using measured data on two different horns in both planar and spherical near-field scanning geometries. The presence and levels of echo pollution in the measurements were controlled by introducing known scattering objects in the anechoic chambers, and comparing with reference situations without disturbance.