Data Recovery Case File · Cameras, Drones & Cards · Refuse Both Prompts
Afraid to Touch It, Which Is Correct
Her enquiry is about primary research data and it ends with the right instinct. Roughly ten hours of video for a doctorate, on a card from a field recorder: "it was working the other day, but now it says SD card invalid, and when I plug it into my computer it says I need to format the card. Losing the data would be disastrous, so I am afraid to touch it." That fear is the correct response and it should be trusted — because both devices are offering her the same destructive remedy in different words, and refusing both is what keeps this recoverable.
| Media | microSD card from a field recorder holding approximately ten hours of research video — reported invalid by the recorder and prompting for a format on a computer |
| Reported situation | Card in normal use in a field recorder · recorder reporting the card as invalid · computer prompting that the card requires formatting · no operations attempted by the owner · approximately ten hours of primary research video held |
| Fault class | Filesystem structures unreadable over intact content — recorder files large, contiguous and signature-identifiable |
| Equipment used | Both format prompts refused and card removed from use · imaged write-blocked (DeepSpar USB Stabilizer 10Gb) · structures rebuilt on the image including surviving backup copies · format-aware carving for recorder containers alongside · footage validated by playback and reported by duration |
The decode: why both devices say the same thing, and why the content survives
Why the recorder calls it invalid and the computer offers a format: the same finding, phrased by two devices with different vocabularies. Both read the card's filing structures, both failed to interpret them, and both offered the only remedy they have. The recorder simply declares the card unusable; the computer offers to write fresh empty structures so it becomes usable again. Usable, in that sentence, means empty — and nothing in either message conveys that.
Why the footage is almost certainly intact: filing structures are small and live in a specific region. Ten hours of video occupies the overwhelming majority of the card, in the data area, entirely untouched by whatever damaged the index. What is damaged is the description, not the recording.
Why recorder footage is unusually good material to work with: field recorders write large files in long continuous passes without fragmenting, and the containers carry distinctive headers. That means two independent routes exist. The filing structures can frequently be rebuilt — filesystems keep backup copies at known positions — returning files with names and timestamps. And where they cannot, the recordings can be identified by their own internal structure and extracted regardless, which on this kind of media works exceptionally well.
What most likely caused it: an interrupted write. A recorder losing power mid-file, a card removed before writing completed, or a battery failing during a long session leaves the structures half-updated with nothing to reconcile them from, since the filesystems used on cards keep no journal.
The one thing that would end this: accepting either prompt. The recorder will offer to format on the next power-up, and the computer will offer every time the card is connected. She should stop connecting it to either.
The two things worth doing today, separately from the recovery: tell the supervisor now, factually, because institutions have procedures for technical loss affecting research and a note in advance is worth far more than an explanation after a deadline. And, once this is resolved, treat the card as the working copy rather than the archive — ten hours of irreplaceable primary data on a single card in a recorder is a research-continuity risk rather than a filing choice, and offloading to two destinations after each session is the standard practice for exactly this reason.
On the bench
Both format prompts were refused and the card removed from use, since the recorder and the computer each offer to write fresh structures over the description of the footage. The card was imaged write-blocked behind the DeepSpar USB Stabilizer 10Gb. Structures were rebuilt on the image from their surviving backup copies, returning files with names and timestamps, with format-aware carving for recorder containers run alongside as an independent route. Footage was validated by playback and reported by duration against the ten hours expected.
The outcome
Both prompts refused, the card imaged and the structures rebuilt with carving run in parallel, validated by playback against the expected duration. 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, for anyone whose recorder and computer both reject a card: they have made the same finding in different words, and both remedies are destructive. Your footage occupies almost all of the card and is untouched; what is damaged is the small region describing it. Recorder files carve exceptionally well, so two independent routes exist. Refuse both prompts.
Research or recording card that both devices reject
Your instinct not to touch it is right — keep doing that. The recorder calling the card invalid and the computer offering to format it are the same finding in two vocabularies: both read the filing structures, neither could interpret them, and each offered the only remedy it has. Formatting writes fresh empty structures over the description of your recordings. Those structures are small and sit in one region, while ten hours of video occupies almost the entire card and is untouched, so what's damaged is the description rather than the footage. Recorder files also carve exceptionally well, being large, continuous and distinctively headed, so there are two independent routes back. Tell your supervisor now, and in future offload to two destinations after each session.
Refuse both prompts — call Newcastle Data Recovery on 0191 406 1051; imaged write-blocked, structures rebuilt from backup copies with carving run in parallel, footage validated against expected duration.
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.