Data Recovery Case File · Solid State & Flash · Two Places, One Answer

Two doors to the same footage, one clearly better: the failing SSD versus the emptied cards weighed honestly — why professional video carves worse than JPEG, and the clips recovered from the source that never truly lost them

His enquiry was a clinic in self-diagnosis. "I have lost .MOV video files. I was copying from SD card to a Seagate SSD when the SSD became unresponsive." The compounding steps, reported honestly: "the SSD had 2 or 3 SD cards just copied over, and I had deleted the files from these, then emptied the bin. Time Machine is not activated." His own recovery attempt, and its revealing result: "I have not formatted the SD cards, and have tried recovering the files — but the clips are blank and white noise plays after a second. I think the video format is complicated; the JPEGs recovered fine." His SSD reading: "SMART status is failing." And the exactly-right question he posed: "perhaps you have a way of fixing the corrupt files on the SD cards, or maybe you can get in the SSD to extract the files? Hopefully one of these is possible." Both fronts are real; this page weighs them, names the stronger, explains why his video carved worse than his stills — and recovers the movies from the door most likely to open.

The decode: why his JPEGs came back and his videos didn't, and which door to open

The white-noise clue, explained: the single most useful detail he provided is that JPEGs recovered fine while the video came back as white noise — and it explains everything about which tool to use. A JPEG is one small, self-contained file, sitting in one place; simple carving finds it whole. Professional camera video is the opposite: large files, often fragmented across the card, wrapped in a complex container format where the index describing how the pieces fit is as vital as the footage itself. Blind carving — grabbing a run of data from the file's start — captures the container's opening and then, at the first fragmentation boundary, veers into unrelated data: which plays as exactly what he saw, a second of video then white noise. His footage isn't destroyed; it's fragmented and needs structure-aware reassembly, not the blunt carving that served his stills. The video's "complexity" he suspected is real, and it's the whole reason his results split by file type.

Two fronts, weighed honestly: he identified both doors correctly; the bench's job is to rank them. The SSD holds the freshly-copied set but is SMART-failing — a drive announcing its own decline, risky to lean on, requiring managed-power imaging before it degrades further. The cards hold the originals: he deleted the files and cleared the bin, but crucially never formatted, and wrote nothing since — so on healthy card media, the clips sit intact-but-unlinked, the most favourable recovery state there is. The verdict: the cards are the stronger, safer front — stable media, unformatted, contents present — with the failing SSD imaged in parallel as corroboration and fallback. Two doors; open the sound one first.

His conduct, graded: not formatting the cards was the decisive good call; stopping his own attempts once the video came back as noise (rather than grinding on) preserved the media; and naming both possible fronts gave triage its map. The one gap — no Time Machine — is noted only as the reason two damaged fronts had to substitute for one intact backup, and as the habit to fix next.

The recovery: the sound door first, the clips reassembled

The source cards were imaged write-blocked, and the video was recovered the way his stills hadn't needed — structure-aware: the container formats parsed, fragmented pieces traced and reordered by their own internal indexing rather than grabbed in blind runs, each clip rebuilt from its scattered parts. The failing SSD was imaged in parallel under managed power, both as insurance and to fill any clip the cards couldn't complete. The .MOV files came back playing — full clips, not a second and static — verified by watching, drawn chiefly from the cards that had never truly lost them.

On the bench

The two fronts got the two right instruments. The unformatted cards — the stronger door — were imaged behind the DeepSpar USB Stabilizer 10Gb and the video rebuilt structure-aware: container indexes parsed, fragmented runs traced and reordered by the format's own internal bookkeeping, the reassembly blind carving cannot perform (and the precise reason his clips had played one second of picture and then noise). The SMART-failing SSD was imaged in parallel on the Atola Insight Forensic under managed power as corroboration and gap-fill. Verification was the only kind that counts for footage: every clip watched — openings, middles, endings — before delivery.

The outcome

The video delivered verified on new media, recovered from the stronger of the two fronts he'd correctly identified, with the failing SSD's image held as corroboration. Free assessment, one fixed written figure including VAT; on cards and sticks where content has been deleted or overwritten, the figure is payable upfront. His question — fix the cards or open the SSD? — answered by doing the smarter version of the first: not "fixing corrupt files," but reassembling intact-but-fragmented ones from unformatted media. And the habit for next time, gently: a backup that's actually switched on turns two damaged doors into none.

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.