Data Recovery Case File · Cameras, Drones & Cards · Right Place, Wrong Edges
A Carved File of the Right Size Can Still Have the Wrong Contents at Both Ends
Her enquiry describes a recovery that produced output without producing results. A formatted card whose video was recovered with consumer software, leaving "a number of large files that I am unable to play." Size is not evidence of correctness: a carving tool finds where a clip begins and estimates where it ends, and a file with the wrong boundaries is exactly this large and exactly this unplayable.
| Media | Camera memory card previously formatted — video content recovered by consumer software into files of substantial size that do not play |
| Reported situation | Memory card formatted · video content recovered using consumer recovery software · recovered files present at substantial sizes · files not playing in any player · owner offering a sample for examination · original card understood to be available |
| Fault class | Carved video with incorrect extents — file openings located and continuation estimated; container structures absent or misaligned |
| Equipment used | Recovered output assessed against the original card rather than repaired in isolation · card imaged write-blocked before any further scanning · stream boundaries derived from internal frame structure rather than estimated length · container structures rebuilt around correctly delimited streams · output validated by playback in full |
The decode: what carving gets right, and what it guesses
What a carving tool does well: finding beginnings. Video files start with a recognisable pattern identifying the format, and locating those across a card is reliable and fast.
What it cannot determine the same way: where a clip ends. There is no equivalent marker for the end of a file, so the tool either reads a length from a header it may not have, or assumes a clip continues until the next recognisable beginning.
Why that assumption produces large files: it sweeps up everything between one opening and the next. If the following clip's opening was not found, the file continues far beyond where it should — which is why the sizes look plausible and generous.
Why the result will not play: a video file is not just a stream but a container describing it. Players read structures giving duration, format and an index of where frames sit — and a carved file with the wrong extent has those describing something that is not there.
Why the format she has makes this particularly likely: some video containers write their descriptive structures at the end of the file rather than the beginning. A clip carved with the wrong ending loses exactly the part that describes it, leaving a stream nothing can interpret.
Why that is a hopeful diagnosis rather than a bleak one: the footage is present in those files. What is wrong is the delimitation and the description, not the content — and both are reconstructible from the stream itself.
How the correct boundaries are found: from the internal structure of the video rather than from any header. Frames follow one another in a predictable pattern, and where that pattern stops is where the clip stops — which is a measurement rather than an estimate.
Why working from her recovered files alone is second best: they are already delimited wrongly, and material may have been omitted between them. Working from the original card allows the whole surface to be reconsidered, so clips missed entirely can also be found.
Why the card must therefore still be untouched: everything depends on it. The recovered files are a partial result; the card is the source, and it should not have been used since.
What she should send: the card, and a sample of the output. The sample identifies the format and the tool's behaviour quickly, which shortens the work.
On the bench
Recovered output was assessed against the original card rather than repaired in isolation — carving reliably locating file openings by recognisable patterns while having no equivalent marker for endings, so extent is read from a header that may be absent or assumed to run to the next opening, producing oversized files. Some video containers place descriptive structures at the end, which a wrong extent omits. Stream boundaries were derived from internal frame structure rather than estimated length.
The outcome
The output assessed against the original card, boundaries derived from internal frame structure, and containers rebuilt around correctly delimited streams. Free assessment, one fixed written figure including VAT; on cards where content has been deleted or overwritten, the figure is payable upfront. The decode: size is not evidence of correctness. Carving finds where a clip begins reliably and estimates where it ends — and some video formats keep their describing structures at the end, which a wrong ending discards.
Recovered video files that will not play
Keep the original card untouched and send it along with a sample of the output — the recovered files are a partial result and the card is the source, so working from it allows clips the tool missed entirely to be found as well. Understand why the files are large and unplayable: carving locates file openings reliably but has no equivalent marker for endings, so it assumes a clip runs to the next opening. That produces oversized files, and some video formats keep their describing structures at the very end, which a wrong ending discards. The footage is usually present inside them.
Keep the original media — call Manchester Data Recovery on 0161 871 0788; output assessed against the source, boundaries derived from internal frame structure, containers rebuilt around correctly delimited streams.
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.