3. Zhang & Wen [1] filtered observed seismograms to short periods, and speculated that some signals may be the SPdKS-SKPdS phases, and they further extrapolate their speculation that those short-period signals may be filtered into long period SKS waves. There is a major flaw in this reasoning. The short-period waves and long-period waves may be related to each other only when there is no other waves arriving at the same time. This is not the case when the seismograms are filtered to short periods. In addition to the weak arrival of the diffracted waves (if any at all), the dominant signals on the short-period seismograms are strong scattered waves generated by shallow heterogeneities (the strength of scattering increases exponentially with frequency). The travel times and strong amplitudes of the scattered waves observed on the short-period seismograms are not correlated with long-period SPdKS-SKPdS waves in any way, as also confirmed by the sensitivity calculations discussed above.