Case file · Memory cards · SDR-2025-0907
The Shoot's Footage Was Corrupt.
Home after a fortnight of filming in Portugal, the memory card came back damaged: three clips on it won't open in anything
. The producer had read the problem accurately and hopefully, they're all still there, right size, just corrupt
, not erased, and the client screening was already in the diary.
Seeing the same thing yourself?
0800 6890668
The translation.
This assumption was correct. A professional video file is a container that includes all of the encoded images with an index or table of contents wrapped around them that lists each frame's location in the file. When you stop writing before you finish (you run out of battery, pull the memory card too quickly, or there isn't enough room on the card) the image data will usually be written to the card but the index will not yet have been completed, because it is always written last. All of the footage exists; the directions to it do not, and no player will open a file it cannot navigate.
Kit used on this job.
What happens in a case →| Which platform | Its role in this case | Why we use it |
|---|---|---|
| Klennet Carver | Reassembled the fragmented clips into files an edit suite would actually accept | Carving that understands video hierarchy, where other tools hand back unusable debris |
| PC-3000 Flash | Read beneath the filesystem, at a level the reported damage never reached | Pulls data straight from memory chips, with a current library that identifies them automatically |
| UFS Explorer Professional Recovery | Pulled the undamaged clips and the folder layout out of the card image | Handles the difficult filesystems better than most — APFS, ReFS, XFS, ZFS, Btrfs |
In the lab.
Copy the card first, deadline or no deadline
The card was imaged before anything else was attempted. A screening date is a fine reason to work quickly and a poor reason to work on the only copy. Everything after that happened on the image, with the card itself set aside untouched in case a second run at it was wanted.
Sort the clips that opened from the clips that didn't
Clips that were intact and healthy were extracted using normal recovery methods and set aside. The three damaged clips were examined very closely, and it was confirmed that the picture data was successfully placed across the card, but the container indexes were either truncated or completely missing. This split is what determines the overall repair process.
Use the shoot's own working clips as the template for the rest
When you are dealing with fragmented cinema video, general purpose tools give up: no editor will import the files they return. As such, healthy clips from the same camera body, shot on the same day with the same codec, were utilised to create a structural template, each broken container was rebuilt around the surviving picture data, and all of the resultant files were validated by being watched through, rather than simply believing the duration reported.
The result.
Two of the three clips returned whole and were able to play from start to finish (visually verified as opposed to checking for file size) prior to showing. The last clip lost the final few seconds of video when the battery died. We retired the card.
Similar cases on the index.
Also filed under Memory cards.
Does this sound like your own drive?
Like with each of the previous examples listed above, the rule applies again: turn it off first, then allow for a free diagnosis of your problem before making any decisions about what you should do next.