Nick Berg decapitation video: analysis, part 1
There are a number of reports circulating that question the authenticity of the Nick Berg decapitation video (Al Jazeera has a summary). Many of these reports are clearly politically motivated, but they do raise several issues with the video – specifically, audio-video sync errors, and the peculiar lack of blood visible following the decapitation.
While there are plenty of opinions being raised as to the implications of these issues, we’ve seen very little objective or verifiable analysis. The following is part one of a series of posts offering some basic analysis of the evidence provided by the video itself. At the end of the series we will present several conclusions about the authenticity of the video and the events depicted within.
First up: identifying the file, since there are various edited copies available. The copy of the video obtained by Vigilant TV is a Microsoft WMV file, 5.30 MB (5,566,007 bytes) in size, with the MD5 checksum f89501168c66439056a835ee3c4a1061; its length is reported as 5:37, 132kbps (the video cuts out at 5:27; the length disparity could be caused by audio sync issues). This appears to be the most commonly available full length version. Several news articles have reported audio sync errors that are consistent with this copy, suggesting this is at least very similar to the one analysed in mainstream reports.
The video has a timestamp in the lower right corner, consistent with the time function often provided by consumer video cameras. There is no date. The timestamp jumps on several occasions; below is a table showing the significant events:
| Elapsed Time | On-screen Timestamp | Description |
|---|---|---|
| 0:10 | jumps from 13:26:27 to 2:18:33 | two shots showing Nick on a chair, with an obvious edit; audio appears more or less in sync at this point |
| 0:20 | jumps from 2:18:43 to 2:40:03 | cut to Nick on the floor, in front of 5 hooded people |
| 4:28 | 2:44:32 | audio of screaming begins; video is clearly unsyncronised |
| 4:37 | 2:44:41 | video of attack begins. Nick can be seen shifting uncomfortably just before he is grabbed, but otherwise doesn’t react |
| 4:39 | jumps from 2:44:42 to 13:45:48 | cut to a closeup of a black-hooded attacker cutting and sawing at Nick’s neck, while a white-hooded assailant holds him down. Note that during this portion of the video, a person is sometimes visible on the right of the screen wearing grey clothing that is lighter than any of the hooded figures visible earlier |
| 5:18 | jumps from 13:46:33 to 13:47:47 | cut to the white-hooded attacker raising Nick’s severed head |
| 5:26 | to 2:46:19 | cuts to a slighly wider shot of the white-hooded attacker holding Nick’s severed head |
At the beginning of the attack, the audio is running approximately 9 seconds ahead of the video.Note that from 0:20 to 4:28 (the speech), 4 minutes and 8 seconds have elapsed, while the video timestamp has incremented by 4 minutes and 29 seconds. There are no significant breaks or obvious edits during this time. This suggests that the video was playing slighly faster than it should have been when the file was encoded. (The audio was running faster still)
Note the disparity of the timestamps: some shown in the format 2:xx:xx, others as 13:xx:xx. There are two possible explanations for this: first, that the events take place over 11 hours or so (2:18am to 1:47pm); or second, that two cameras were used. The grey-clothed figure visible during the decapitation sequence may have been a second camera operator.
Coming next: an analysis of the two camera hypothesis.