@Jiayu Lou We understand you wish to see how the macro {Best} attribute works to select the thumbnail. Please see the below response from the product group:
The algorithm to pick the {Best} thumbnail is quite complicated. Basically, it analyzes the histograms of those frames in range and then picks the closest frame to the center of the histogram clustering. As an intuitive example, if there are 101 frames in total and 51 of them are black images while 50 are white images, this algorithm will pick a black image because the histogram of a black image is the closest to the center of all black and white images. In other words, it represents the majority of those 101 images. Hope this helps.
If this answer was helpful, please mark it as the answer to help others who may have the same question. If you have further questions or concerns, please let us know by replying to this message so we can assist you further.