Data Recovery Case File · Cameras, Drones & Cards · Failure at Capacity
A Card That Fails as It Fills May Never Have Held the Capacity It Claimed
His enquiry pairs a failure with the moment it happened, and the pairing is the diagnosis. A large card that "recorded video all day, and as it filled up it stopped working" — now unrecognised in any camera, unmountable, and absent even from the system's disk utility. Failure correlated with approaching capacity is a specific signal, and one of its explanations is that the card was never the size it reported.
| Media | 256GB high-speed memory card — failing during extended recording as capacity was approached; not enumerating on cameras or hosts |
| Reported situation | Card used for video recording throughout a day · failure occurring as the card approached capacity · card not recognised in any camera · card not mounting on a computer · card not appearing in the system's storage utility · content required |
| Fault class | Failure correlated with capacity — controller fault under exhausted spare blocks, or reported capacity exceeding physical memory; content extent determined by actual capacity |
| Equipment used | Correlation with capacity treated as diagnostic rather than coincidental · physical memory capacity established directly rather than from the controller's report · card read on fixed contact equipment independent of host enumeration · memory read past the controller with translation layer reconstructed in software · recovered footage validated by playback in full |
The decode: two explanations, and why the timing points at one
Why the moment of failure is informative: cards do not usually fail on a schedule. A fault arriving precisely as capacity is approached is correlated with something the card only does when full.
The first explanation — spare blocks exhausted. Cards hold reserve memory to replace blocks that wear out, and the controller draws on it silently. A card near capacity has both less free space and, if it is old, less reserve, and a controller unable to complete a write may fail entirely.
Why that produces total failure rather than a full-card message: the controller's own management structures live in the same memory. A controller that cannot write its tables cannot start, which is why nothing sees the card at all.
The second explanation, and the timing points at it strongly: the card may never have held what it claimed. Cards exist whose controllers are configured to report a capacity far larger than the memory fitted, and they behave perfectly until the real capacity is passed.
Why that produces exactly this symptom: writing beyond the physical memory has nowhere to go. Addresses wrap around and overwrite earlier content, or the controller fails outright — and either way it happens at a predictable point.
Why an all-day recording is precisely the test that finds it: most use never reaches capacity. A card can be used for years without approaching its claimed size, and the fault stays hidden until something fills it.
How the two are distinguished: by establishing the physical memory directly. The controller's claim is not consulted; the packages themselves are read and their true capacity determined.
Why that determines what can be expected: under the first explanation the whole recording may be present. Under the second, only what fitted within the real capacity exists at all, and the rest was never stored.
Why saying that in advance matters: it sets a realistic ceiling. An owner expecting a full day and receiving whatever fitted should know why before the work rather than after.
What must not happen: no further insertion into any camera. A card that fails to enumerate invites a format offer, and reinsertion works contacts that may already be part of the problem.
On the bench
The correlation with capacity was treated as diagnostic rather than coincidental — cards holding reserve memory drawn on silently as blocks wear, so a card near capacity with depleted reserve may leave the controller unable to write its own management structures and therefore unable to start. Alternatively, controllers configured to report capacity exceeding the memory fitted behave normally until the real capacity is passed. Physical memory capacity was established directly rather than from the controller's report.
The outcome
Correlation with capacity treated as diagnostic, physical capacity established directly rather than from the controller's claim, and memory read past the controller. 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: failing exactly as it filled is the diagnosis. Either the reserve blocks ran out, or the card never held the capacity it claimed — and that determines how much of the day exists at all.
A card that fails at the point it fills up
Stop inserting it into cameras, since a card that won't enumerate attracts a format offer and each reinsertion works contacts that may be part of the problem. The timing is the useful part: a fault arriving exactly as capacity is approached points at something the card only does when full. Either its reserve blocks are exhausted and the controller can no longer write its own tables, or the card was reporting a capacity larger than the memory actually fitted — which behaves perfectly until the real limit is passed. Ask for the physical capacity to be measured.
Keep it out of the camera — call Manchester Data Recovery on 0161 871 0788; correlation with capacity treated as diagnostic, physical capacity established directly, memory read past the controller.
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.