00f11ed9-afdc-411e-9938-7bf3886360a5

It's intriguing that the tools aren't recognizing these videos as duplicates, especially when you're sure they are. The issue likely stems from the different encoding formats and bit rates, which can affect the hashing methods used by the tools.

To address this, you can use a more advanced method that compares the actual video content rather than just the file metadata and hashes. One approach is to use the Structural Similarity Index (SSIM) or Perceptual Hash (pHash) methods, which focus on the visual content of the videos.

Here’s how you can use ffmpeg and OpenCV to compare the visual similarity between two videos using SSIM in Python