Spectral analysis of heart rate variability (HRV) is an accepted method for assessment of cardiac autonomic function and its relationship to numerous disorders and diseases. Various non-parametric methods for HRV estimation have been developed. The spectrum of counts, the instantaneous heart rate spectrum and the interval spectrum are mostly practised. Although extensive literature on their respective properties is available, there seems to be a need for a more complete comparison, eventually resulting in recommendations for applicability. The methods for HRV spectral analysis use their specific transforms of the primary R-R interval series into input signals for spectral computation. This is, in fact, the reason for obtaining different spectra. A basis for comparison is established, revealing the generic relationships of these HRV input signals. It allows for a more systematic evaluation and for further development of the considered methods. The results with simulated and real HRV data show better performance by the spectrum of counts and by a proposed instantaneous heart rate spectrum, obtained using a cubic spline interpolated input signal. The modulation depth of the primary signal can influence the accuracy of the HRV analysis methods.