This study delves into the impact of voids within CFRP materials on dark-field images obtained through an X-ray Talbot-Lau interferometer (TLI). Detecting voids in CFRP holds excellent significance because these voids can significantly influence material properties, such as strength and elastic modulus. TLI presents a promising approach for void detection, given its ability to capture a wide area rapidly and effectively identify voids. However, the relationship between voids and dark-field images has been inadequately assessed. Quantitative assessment of the impact of voids on the dark-field signal is imperative to predict the characteristics of voids in CFRP based on dark-field images. This study introduced unimpregnated areas, representative of voids in CFRP, within the material. Subsequently, we photographed the CFRP using TLI and verified that these unimpregnated areas influenced the dark-field signal. This result was used to thoroughly investigate the relationship between the geometric properties of unimpregnated areas and the extent of their impact on the dark-field signal. Our findings revealed a correlation between the geometric properties of these void-like areas and the magnitude of their effect on the dark-field signal.