Vibrations of engineering structures can give insights into their dynamic properties and aid in assessing their health conditions and identifying damage. One-way scanning laser Doppler vibrometer (LDV) aims to scan structures along a certain path without stopping. The signal quality of one-way scanning LDV is affected by the surface characteristics of target structures. Different materials of engineering structures, such as steel, clay, and asphalt, exhibit different textures, roughness, and particle sizes. These differences can cause variations in backscattering and speckle patterns along the scanning path, affecting signal quality. This paper investigates how different materials and scanning speeds affect the signal quality of one-way scanning LDV through experiments. A rotating mirror directs the laser beam of an LDV to scan a vibrating beam with target surfaces made from clayey soil, sandy soil, steel, asphalt and wood at two different speeds: 0.6 ms^-1 and 3 ms^-1. Subsequently, a two-step despeckling algorithm involving moving root mean square-based thresholding and an Empirical Mode Decomposition-based filter is applied to separate out the noise from the signal. The results indicate that noise power is much higher and signal-to-noise ratio (SNR) is significantly lower for clayey soil, sandy soil, and asphalt compared to wood and steel.