Palm vein-based biometric identification offers a higher level of security than traditional methods such as fingerprinting, iris, or facial recognition. One of its main advantages lies in using internal body features, which makes it highly secure and less susceptible to external changes. However, its large-scale application is limited by the need for large-scale public databases. In this context, synthetic palm vein image databases partially address this challenge, as there will always be a difference between synthetic and real. To mitigate these gaps, we propose to evaluate the differences using a spectral perspective and present techniques to fit the magnitude spectrum and power spectral distribution. We evaluated the similarity of the resulting synthetic images to the real images from the most representative state-of-the-art palm vein databases. The proposed approaches help to reduce the difference between synthetic and real images from the CASIA database, improving the accuracy in the representation of synthetic palm veins for the evaluation of biometric recognition algorithms.