General Sampling Expansion Reconstruction Method (GSERM) and Digital Spectrum Reconstruction Method (DSRM), which prove effective to reconstruct azimuth signal of Displaced Phase Center Apertures (DPCA) Synthetic Aperture Radar (SAR) system from its Periodic Non-Uniform Sampling (PNUS) data sequences, would amplify the noise and sidelobe clutter simultaneously in the reconstruction. This paper formulates the relation of the system transfer matrixes of the above two methods, gives the properties, such as periodicity, symmetry, and time-shift property, of their Noise and Sidelobe Clutter Amplification Factor (NSCAF), and discovers that DSRM is more sensitive than GSERM in the white noise environment. In addition, criteria based on initial sampling point analysis for the robust PRF selection are suggested. Computer simulation results support these con-clusions. General Sampling Expansion Reconstruction Method (GSERM) and Digital Spectrum Reconstruction Method (DSRM), which prove effective to reconstruct azimuth signal of Displaced Phase Center Apertures (DPCA) Synthetic Aperture Radar (SAR) system from its Periodic Non-Uniform Sampling sequences, would amplify the noise and sidelobe clutter simultaneously in the reconstruction. This paper formulates the relation of the system of transfer matrixes of the above two methods, gives the properties, such as periodicity, symmetry, and time-shift property, of their Noise and Sidelobe Clutter Amplification Factor (NSCAF), and discovers that DSRM is more sensitive than GSERM in the white noise environment. In addition, criteria based on initial sampling point analysis for the robust PRF selection are suggested. Computer simulation results in these con-clusions.