A system for extracting a video-watermark embedded in a video sequence, comprising a decryption and decoding engine to receive the video sequence in a video-codec-encoded format and encrypted, then decrypt the video sequence yielding a decrypted video sequence, then decode the decrypted video sequence yielding a decrypted decoded video sequence, and then output the decrypted decoded video sequence to a display device for rendering thereon, a measurement module to measure memory usage utilized by the engine decoding at least part of the video sequence yielding measurements of memory usage, a memory usage analyzer to analyze the measurements to determine an arrangement of video frames of at least one frame type in the video sequence, and an encoding feature analyzer to analyze the arrangement of the video frames to extract the video-watermark from the video sequence. Related apparatus and methods are also described.