Data Recovery Case File · Cameras, Drones & Cards · His Observation Is the Evidence
Full-Size Files That Play as Nothing Are Missing Their Index
His enquiry contains the observation that settles it, offered as a hopeful aside. Footage recorded abroad that he tried to transfer: "all of the files just seemed blank — they have size to them, so there is something there, but I think they might be" damaged. He is right, and the reasoning is sound. A file that reports a real size and plays as nothing is a file whose content exists and whose map does not.
| Media | Action camera media — video files present at full recorded length; content not playable; file sizes consistent with recorded footage |
| Reported situation | Video footage recorded on an action camera during travel · files transferred from the camera media · files appearing empty when opened · file sizes substantial and consistent with recorded content · footage required |
| Fault class | Video index absent or incomplete — recording terminated before trailing metadata was written; frame data present and contiguous |
| Equipment used | File length reconciled against occupied capacity before any conclusion · files examined structurally rather than by playback · index rebuilt by scanning for frame boundaries · reference recording from the same device and settings used to supply encoding parameters · output validated by playback |
The decode: why the size is the whole finding, and how video files are built
Why his reasoning is correct: a file that was never written has a length of zero. A file reporting hundreds of megabytes contains hundreds of megabytes of something — so the content reached the medium, and whatever is wrong is downstream of that.
How most video formats are arranged: a long run of compressed frame data, followed by an index describing it — where each segment begins, what the timings are, what the encoding parameters were. The index is written at the end, once recording finishes, because until then the camera does not know how long the recording will be.
What follows if recording does not finish cleanly: battery exhaustion, a card removed, a crash, or a camera switched off abruptly. The frame data is all there and the index was never written, so every player opens the file, looks for the map, finds nothing, and reports it as unplayable or empty.
Why that is a very good position: the footage is complete. Nothing is missing except the description of it, and a description can be rebuilt — by scanning the file for the boundaries between segments and reconstructing the index from what is actually there.
Why a second file from the same camera helps enormously: a recording that completed normally contains a valid index made by the same device at the same settings. That supplies the encoding parameters directly, which turns a reconstruction from an inference into a transfer. Anybody with damaged footage should keep an intact clip from the same camera for exactly this reason.
What to check before assuming anything, and it is free: open one of the files in something that examines rather than plays. A media tool that reports on a file's contents will say whether frame data is present, which distinguishes a missing index from genuinely empty space.
What must not happen: no reformatting of the card, and no further recording on it. If the files were transferred but the originals remain on the card, those originals are the better source — a transfer of a damaged file produces a damaged copy, and the card may hold data the transfer did not capture.
Why action cameras produce this constantly: they are used until the battery dies, in conditions where recordings are stopped by circumstance rather than by a button.
On the bench
File length was reconciled against occupied capacity before any conclusion — a file reporting substantial size containing substantial content, so the fault is downstream of the write. Files were examined structurally rather than by playback, most video formats storing frame data followed by a trailing index written only when recording completes, so an interrupted recording leaves complete footage with no map. The index was rebuilt by scanning for frame boundaries, with a reference recording from the same device and settings supplying encoding parameters.
The outcome
File length reconciled against capacity, files examined structurally rather than by playback, and the index rebuilt using a reference recording for parameters. Free assessment, one fixed written figure including VAT; where a chip has to be removed, 50% of parts and labour is payable upfront with the balance only on success — otherwise no recovery, no fee. The decode: your observation is the finding. A file that was never written has zero length — yours report real sizes, so the footage is there. Video formats write their index at the end, once recording completes.
Video files that transfer but play as nothing
Keep an intact clip from the same camera at the same settings — that's the single most useful thing you can supply, because a working recording contains a valid index made by the same device, which supplies the encoding parameters directly rather than having them inferred. Your own observation is the diagnosis: a file that was never written has zero length, so files reporting real sizes contain real footage. Most video formats store frame data followed by an index written only when recording finishes, so a battery dying or a card being pulled leaves complete footage with no map, and every player refuses it. Don't reformat the card or record over it.
Keep an intact clip from the same camera — call Oxford Data Recovery on 01865 593000; length reconciled against capacity, files examined structurally rather than by playback, index rebuilt from frame boundaries.
Request a quote online →
Our case files are drawn from genuine enquiries received by our laboratory over the past ten years, anonymised to protect client confidentiality. Each one describes the diagnostic and recovery procedure our engineers apply to that fault, using the equipment listed.