Information loss in video tends to reduce its visual quality since some glitches or artifacts can be detected by humans. Over the past few years, great efforts have been made to reduce packet loss and its visibility in videos. Forward error correction (FEC) methods can be used at the encoder to prevent packet loss. Error concealment methods can be used at the decoder to reduce the visibility of lost packets if the error correction algorithms have failed. One way to prevent the decrease of quality in videos due to packet loss is to tag packets at the encoder according to their priority levels. A problem with this prioritization method is that typically the encoder has no information about what kind of error concealment algorithm the decoder uses. This thesis presents the experiment that we performed in order to determine to what extent the encoder prediction of packet importance is independent of the error concealment method used at the decoder. In this experiment, four different types of packet loss were introduced into nine different videos. Each lossy version of each video was then decoded using five types of error concealment methods. A Visual Quality Metric was then used to rate the concealed videos. Results show that for each type of loss, the error concealment methods did not differ significantly, thus indicating that it is possible to create a packet priority tagging algorithm that is largely independent of the error concealment method used at the decoder
