In this paper, we analyze and compare the partial discharge (PD) signals from acoustic emission (AE) and high-frequency current transformer (HFCT) measurements with Hilbert-Huang transform (HHT). Commonly, the Fourier transform (FT) method is used to illustrate information in frequency domain. The HHT, a time-frequency method, provides another viewpoint to interpret the PD signals. This method allows for collecting information on PD signals both with regards to the time and frequency domains simultaneously. It has been verified that the HHT can provide more information when compared with FT method. In this work, the marginal spectrum simplifies the results of HHT method fatherly. From the marginal spectrum, the abiding and weak noise source can be properly estimated, and the instantaneous variation can be emphasized. The results reported here show that the HHT method has great promise for cast-resin dry-type transformer PD signals analysis.