Data Recovery Case File · Solid State & Flash · The First Sustained Write Is the Test
It Worked Until It Was Asked to Do Real Work
Her enquiry pairs a recent purchase with a deadline. A drive bought online "that had good reviews. I was using it to edit footage for my graduation project, and as I was creating proxies it ejected itself and is no longer showing on any computer." Failing during proxy generation is not a coincidence of timing — that is the first genuinely heavy sustained write the drive had ever been given, and a device that fails at exactly that moment has a recognisable signature.
| Media | Recently purchased solid-state drive from an online marketplace — disconnecting during sustained write activity; not enumerating on any host subsequently |
| Reported situation | Drive purchased recently through an online marketplace · used for video editing · failing during generation of proxy media · disconnecting itself during that operation · not appearing on any computer since · project footage required to a deadline |
| Fault class | Failure during first sustained write — controller or memory capability to be verified against declared capacity; misrepresented device to be excluded |
| Equipment used | Declared capacity verified against addressable memory before any conclusion · original source footage identified as a possible alternative before any work · enumeration state read under strict timeouts · chip-level read past the controller where indicated |
The decode: what a first heavy write reveals
Why editing exposes storage that ordinary use does not: generating proxy media reads large files and writes large files, continuously, for a long stretch. Most everyday use never asks for that — documents, browsing and photographs touch a tiny fraction of a drive at low intensity. So a drive can appear perfect for weeks and meet its first real demand the day it is used seriously.
The first explanation, and the ordinary one: early-life failure. Manufacturing variation shows up in the first weeks, which is why devices fail disproportionately when new and why warranties exist. Nothing she did caused it, and a sustained write simply found it.
The second explanation, which is worth naming because of where it was bought: online marketplaces carry a real proportion of misrepresented storage — devices whose controllers report a larger capacity than the memory actually installed, or built with low-grade memory sold as premium. These behave perfectly until asked to write more than they really hold, at which point they fail outright, or silently wrap around and overwrite what was written earlier.
Why that matters beyond this case: the signature is exactly what she describes — a new device, good reviews, fine in light use, and a failure on the first sustained write. Reviews do not detect it, because most reviewers never fill the device either.
What to do with any new storage, and it costs nothing: verify it before trusting it. Free tools write a full pass across the whole declared capacity and read it back, confirming that what was written returns correctly. An hour before you rely on something is worth considerably more than an hour afterwards, and it catches both misrepresentation and early-life defects.
The urgent practical point for her deadline: proxies are derived files, generated from source footage. If the original media still exists — on camera cards, on another drive, or wherever it was first offloaded — then nothing irreplaceable is on that drive at all, and the project can continue while the drive is assessed. That question should be answered today rather than after the recovery.
What the fault is likely to be: a drive that disconnected during writing and now enumerates nowhere has failed at controller level, which on solid-state media places the mapping and often the encryption in the same place as the failure.
On the bench
Declared capacity was verified against addressable memory before any conclusion — misrepresented devices reporting a capacity larger than the memory installed and behaving normally until asked to write beyond it, which produces exactly this failure signature on a first sustained write. Original source footage was identified as a possible alternative before any work, proxy media being derived rather than original. Enumeration state was read under strict timeouts, with chip-level reading past the controller where indicated.
The outcome
The declared capacity verified, alternative sources identified before any work and the memory read past the controller where required. 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: proxy generation was the first sustained write that drive had ever been given, which is when both early-life defects and misrepresented capacity reveal themselves. And proxies are derived — check whether your source footage still exists before anything else.
New drive that failed during heavy editing
Check whether your source footage still exists first — proxies are derived files, so if the originals are on camera cards or another drive, nothing irreplaceable is on the failed device and your project can continue today. Then know why the timing isn't coincidence: generating proxies is the first genuinely sustained write most drives ever receive, and everyday use never asks for anything like it. That reveals early-life defects, and it also reveals misrepresented storage, which marketplaces carry a real proportion of — controllers reporting more capacity than the memory installed, fine until asked to write beyond it. Reviews don't catch that because reviewers don't fill drives either. Verify new storage with a full write-and-read pass.
Check your source footage — then call Guildford Data Recovery on 01483 901310; declared capacity verified against addressable memory, enumeration read under strict timeouts, 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.