Data Recovery Case File · Trust, Practice & Honest Limits · What ASAP Can Mean
Capture Time on Four Terabytes Is Set by the Drive's Condition, Not by the Deadline
This enquiry from a production company states its constraint plainly. An external solid-state drive that failed after a shoot and "won't seem to mount", with around four terabytes to recover and a request to have it "assessed and hopefully recovered as soon as possible." The honest answer separates two timescales: how quickly it can be assessed, which is fast, and how quickly four terabytes can be captured, which the drive decides rather than anyone else.
| Media | External solid-state drive of approximately 4TB — not mounting; commercial production content held; expedited handling requested |
| Reported situation | External solid-state drive used during a commercial production · drive failing after the shoot · drive not mounting on the host · approximately 4TB of content held · assessment and recovery requested at the earliest opportunity · enquiry made on behalf of a production company |
| Fault class | Solid-state device not presenting — controller or interface fault indicated; capture duration governed by read behaviour across a large capacity rather than by scheduling |
| Equipment used | Assessment timescale separated from capture timescale before any commitment was given · device addressed with imposed timeouts outside any operating system storage stack · memory read past the controller where the device did not present · imaged with per-region read verification and progress retained across sessions · content prioritised by production requirement where a partial delivery was useful |
The decode: the two timescales, and where priority actually helps
Why the assessment is genuinely quick: it establishes condition rather than reading content. Whether the device responds, and how it behaves when addressed directly, is determined in hours.
Why the capture is a different matter: four terabytes is four terabytes. Even a device reading perfectly takes many hours to copy at that capacity, and one reading imperfectly takes considerably longer.
Why that duration cannot be shortened by urgency: it is set by the device. Reading faster than the device will return data is not available at any price.
Why saying so plainly is more useful than agreeing: a production can plan around a real figure. A commitment that cannot be met is worse than a longer one that can.
Where speed genuinely is available, and it is worth using: ordering. Where a subset is needed first, the regions holding it are located and captured before the rest.
Why that suits production work particularly: not everything is equally urgent. Selected shots, a specific day's material or a single camera's output can be delivered while the remainder continues.
What has to be supplied for that to work: a specification. Folder names, dates or camera identifiers let the wanted material be located from the structures before bulk capture begins.
Now the fault, and why a solid-state device changes the picture: the failure modes differ. A device not presenting usually has a controller that will not initialise, with the memory behind it intact.
Why that is comparatively good news for capacity this large: reading is not degraded. Where the memory is sound, capture proceeds at full rate once the controller is bypassed, rather than crawling through marginal regions.
What must not happen meanwhile: no further mount attempts. Repeated connection of a device whose controller is failing risks the mapping that a reconstruction depends on.
On the bench
The assessment timescale was separated from the capture timescale before any commitment was given — condition being determinable in hours while four terabytes takes many hours to copy even from a device reading perfectly, a duration set by the device rather than by scheduling. Content was prioritised by production requirement where partial delivery was useful, wanted regions being located from the structures before bulk capture. Non-presenting solid-state devices typically retain sound memory behind a controller that will not initialise.
The outcome
Assessment and capture timescales separated before any commitment, memory read past the controller, and content prioritised by production requirement. 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. The decode: the assessment is quick and the capture is not, and no amount of urgency changes the second. What does help is telling us which material you need first.
Recovery against a production deadline
Separate the two questions when you brief anyone: how soon it can be assessed, which is hours, and how long four terabytes takes to capture, which the device decides. Reading faster than it will return data isn't purchasable. What genuinely buys you time is priority — supply folder names, shoot dates or camera identifiers so the material you need first can be located from the structures and delivered while the rest continues. Stop attempting to mount it meanwhile.
Tell us what you need first — call Cambridge Data Recovery on 01223 655015; assessment separated from capture, memory read past the controller, wanted material located and delivered ahead of the remainder.
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.