Data Recovery Case File · Formatted & Logical Faults · Reformatting Is Not Conversion
A Filesystem Is Not a Setting You Can Change
Her enquiry describes an operation whose name conceals what it does. Trying "to switch formats of the hard drive to make it compatible with my" tablet, and losing access to the contents. Nothing about a filesystem can be switched. It is the entire structure describing where everything on the drive lives — so changing it means destroying one description and building another, and there is no version of that operation which preserves what was there.
| Media | External hard drive undergoing a filesystem change for cross-device compatibility — previous filesystem replaced; contents no longer accessible |
| Reported situation | Drive holding existing content · reformatting undertaken to achieve compatibility with a tablet · previous contents no longer accessible afterwards · recovery of original files sought |
| Fault class | Filesystem replacement with content unwritten over — extent of the operation determinative of what survives |
| Equipment used | Format type and completion state established before any conclusion · post-format write extent measured against capacity · imaged write-blocked before any reconstruction · previous filesystem structures located from surviving copies · files validated by opening |
The decode: what the word conceals, and what survives anyway
What a filesystem actually is: not a property of the drive, like its capacity or its speed. It is a set of structures written onto the drive that record what files exist, what they are called, and which regions of the surface hold each one. Everything you can see on a drive, you can see because those structures describe it.
So what "changing the format" does: writes a different set of structures, built to a different design, describing an empty drive. The previous description is not translated — it is replaced. There is no conversion step, because the two designs record different things in different ways and nothing in the operation attempts to carry anything across.
Why the interfaces make this so easy to get wrong: the filesystem appears as a choice in a dropdown, alongside a name for the drive, exactly as though it were a preference. Nothing about the presentation conveys that selecting a different entry discards everything, and the warning, where there is one, is worded as a formality about erasing the disk.
Why she was doing something entirely reasonable: a drive formatted for one system frequently cannot be written to by another, and reformatting genuinely is the way to make it compatible. The advice she followed was correct and it assumed the drive was empty — which is the assumption that never gets stated.
Why the position is nevertheless good: writing a new filesystem writes a small quantity of structural data and marks the rest of the drive as available. It does not visit the files, which sit exactly where they were. Only subsequent writing removes them, and on a drive reformatted and then noticed, very little follows.
What decides how much survives: whether the operation was a quick one, which writes structures only, or a full one, which writes across the entire surface. Those are very different outcomes and the distinction is worth establishing rather than assuming — a full format that ran to completion is a materially worse position.
What must not happen now: nothing further written to the drive, and no use of it with either device. Every file copied onto it now lands in the space holding what she wants back.
The rule that follows, without exception: copy everything off before changing any drive's format. There is no safe version and no partial version.
On the bench
Format type and completion state were established before any conclusion — a quick operation writing structures only while a full one writes across the entire surface, which are materially different outcomes. Post-format write extent was measured against capacity, only subsequent writing removing content. The drive was imaged write-blocked before any reconstruction, and previous filesystem structures located from their surviving copies, filesystems maintaining duplicates at known positions which a replacement does not always overwrite.
The outcome
The format type established, the written extent measured and the previous structures recovered from their surviving copies. Free assessment, one fixed written figure including VAT; where a drive has to be opened, 50% of parts and labour is payable upfront with the balance only on success — otherwise no recovery, no fee. The decode: a filesystem is not a setting. It is the set of structures recording what exists on the drive and where — so changing it replaces that description rather than translating it, and no conversion is attempted. Your files were not visited, only the description of them.
Reformatting a drive for a different device
Copy everything off first, without exception — there's no safe version of this operation. A filesystem isn't a property of the drive like its capacity; it's a set of structures written onto it recording what files exist and which regions hold each one. Changing format writes a different set of structures describing an empty drive, and nothing translates the old description across. The interfaces make it look like a preference in a dropdown, which is exactly why this happens. If it's already done, stop using the drive with either device: writing a new filesystem doesn't visit your files, so they're still there, and only what you write now removes them.
Stop using it — call Guildford Data Recovery on 01483 901310; format type and completion established, written extent measured against capacity, previous structures recovered from their surviving copies.
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.