Data Recovery Case File · Cameras, Drones & Cards · The Test Was the Damage
Putting the Card Back to Check Is the One Action That Converts Missing Into Gone
Her enquiry records the sequence honestly, including the part that hurt. Four clips recorded on a flight, only two visible on the computer, so she "returned the card to the drone to see if I could see all the videos", recorded a short test — "and the strange thing is that this new video substituted one of the old ones." That is not strange, it is the mechanism: the space the missing clips occupied is exactly where a new recording goes.
| Media | Memory card from an aerial camera — two of four recorded clips not presenting; one further clip overwritten by a subsequent test recording |
| Reported situation | Four clips recorded during a flight · only two clips visible when the card was read on a computer · card returned to the aircraft to verify · short test recording made · new recording found to have replaced one of the previously recorded clips · recovery of the remaining missing material sought |
| Fault class | Clips written without finalisation and unreferenced by the directory — one further clip displaced by subsequent recording; remaining material retained pending further writing |
| Equipment used | Absent directory references distinguished from absent content · card removed from the aircraft as the immediate priority · imaged write-blocked before any scanning · video located by signature and extents rebuilt from stream structure · displaced regions identified and excluded from expectations before any figure was given |
The decode: why two were missing, and why the test cost one of them
Why clips can be recorded and not listed: the directory entry is written last. Video is streamed to the card as it is captured and the entry describing it is completed when recording stops, along with the closing structure inside the file.
What interrupts that: anything ending a recording abruptly. Loss of signal, low battery, a landing that cuts power, or the aircraft powering down mid-capture — each leaves video written and undescribed.
Why the computer showed only two: it lists what the directory describes. The two finalised clips have entries and the other two do not, so they are invisible while being entirely present.
Why the card reported free space it did not have: unreferenced regions count as available. The card's own accounting has no record that those clips exist, so it offers that space to the next recording.
What her test recording therefore did: took the first available space. Which was the space holding an unlisted clip, and the new video was written directly over it.
Why the substitution she describes is the giveaway: a new file appearing in place of an old one is overwriting, seen from outside. It confirms the missing clips were physically present until that moment.
Why that is worth saying rather than passing over: the observation is genuinely useful. It establishes that the remaining missing clip was there too, and that it is recoverable if nothing more is written.
Why the instinct to check was so natural: the aircraft displayed the clips before and the computer did not. Returning the card to the device that recorded them is the obvious way to compare, and nothing warns that recording anything is destructive.
What must not happen now, and it is the whole of the remaining advice: the card stays out of the aircraft. One clip has been displaced and one has not, and only further recording would change that.
How the surviving clip is recovered: located by its own signature and its extent rebuilt from the structure of the video stream rather than from any directory. An unfinalised clip is reassembled from what the stream itself describes.
On the bench
Absent directory references were distinguished from absent content — video streamed to the card during capture with the directory entry and closing structure written when recording stops, so interruption by signal loss, power removal or abrupt landing leaves clips written and undescribed, invisible to a host while physically present. Unreferenced regions count as free space and are offered to the next recording. Extents were rebuilt from stream structure rather than from any directory.
The outcome
Absent references distinguished from absent content, the card removed from the aircraft, and the surviving clip located by signature with its extent rebuilt from stream structure. Free assessment, one fixed written figure including VAT; on cards where content has been overwritten, the figure is payable upfront. The decode: nothing strange happened. Your unlisted clips occupied space the card believed was free, so the test recording was written straight into it — which also proves they were there.
When clips are missing after a flight
Take the card out and leave it out — don't put it back in the aircraft to check, because that's the action that turns missing into gone. Clips are streamed to the card during capture and the entry describing them is written when recording stops, so anything ending a flight abruptly leaves video present but unlisted. The card counts that space as free and offers it to the next recording. Checking on the device means recording, and a few seconds of test footage lands directly on top of what you're looking for.
Keep the card out of the aircraft — call Manchester Data Recovery on 0161 871 0788; absent references distinguished from absent content, imaged before scanning, extents rebuilt from stream structure.
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.