Video copy detection is a crucial computer vision task, essential for protecting intellectual property and identifying reused content across platforms. The rise of social media and short video platforms has made this task more challenging due to diverse video edits and transformations. Current methods typically extract one frame per second, prioritizing accuracy but often lacking efficiency, especially with large video datasets. In contrast, our study takes a bold step by investigating significantly longer intervals for frame sampling. Surprisingly, we found that using much longer intervals does not substantially compromise performance and, in some cases, even improves it by up to 3%. Moreover, this approach can achieve a remarkable speedup of up to 100 times. Our analysis includes a detailed examination of how varying sampling intervals influence the identification of true positives and the occurrence of false negatives. Additionally, we explore how the length of copied segments affects detection performance, assessing the accuracy and reliability of detection methods for shorter and longer segments. The findings of this research highlight the potential benefits of adopting larger sampling intervals, which could lead to significant improvements in processing speed while maintaining or even enhancing detection performance.