Data Recovery Case File · Cameras, Drones & Cards · Entries Intact, Content Missing

Photos Open Blank While the File List Looks Perfect

His enquiry contained an unusually careful observation and a conclusion that is nearly right. A camera microSD card with "610 photos, all approximately 3MB each. I can open the directory and see all files with associated dates created, which are the dates when each picture was taken. However when I try to open an image it is blank, as if there is no data there at all. The images also don't show correctly on the camera preview. Explorer freezes when I try to copy them. I think the data for the images is still present but corrupt." His diagnosis needs one refinement, and the refinement is the good news. The data is not corrupt. It is unreadable — and those are different things, because corrupt data has been altered and unreadable data is simply not being returned. This page explains why a perfect file listing can sit over content that will not come back, and why patient re-reading is the answer rather than repair.

The decode: two reads, and why the listing survives

Why the listing is perfect and the pictures are not: a card holds two very different kinds of information in two very different places. The directory — filenames, sizes, dates, and pointers to where each image begins — is small and stored in a concentrated area near the front. The image content is spread across the rest of the card, wherever there was room when each photograph was written. Those are separate reads. If the directory region still returns data cleanly and the wider area does not, the result is exactly what he describes: a complete, accurate, entirely convincing list of 610 photographs with correct dates and sizes, and nothing behind any of them. The listing is not lying; it is reporting what it read.

Why "unreadable" is better than "corrupt": the refinement to his conclusion. Corrupt data would mean the stored content had been altered — overwritten, scrambled, damaged in place. What is happening here is that the card is failing to return what is stored, which is a reading problem rather than a content problem. That distinction matters because unreadable sectors are frequently recoverable: a sector that fails on one attempt often succeeds on a patient repeat under proper timeout control. Corrupt content would not.

Why everything freezes: his second observation confirms the first. File managers and cameras have no way to limit how long they wait for a card that does not answer — so faced with sectors that never return, they stall indefinitely, taking the application or the device with them. The freezing is not an incompatibility; it is the card refusing to answer, made visible. And it means every copy attempt is a prolonged session of a struggling card being hammered on exactly its weakest regions.

Why the camera shows them wrongly too: useful corroboration. If the fault were on his computer — a bad reader, a driver problem — the camera would display the images normally. That both devices fail identically places the fault in the card, and rules out everything external in one observation.

What actually recovers them: one imaging pass with per-sector timeouts, so a failing sector is budgeted and deferred rather than retried until something hangs; the healthy expanse captured first at full speed; and the weak regions revisited afterwards on their own passes, where repeated patient reads recover data that a single attempt cannot. The images are then rebuilt from sectors genuinely read, and validated by rendering rather than by checking that a file of the right size exists.

On the bench

Copy attempts stopped, which was the most valuable single step available. The card was imaged write-blocked behind the DeepSpar USB Stabilizer 10Gb with per-sector timeouts enforced — the decisive difference from what his file manager had been doing — and the sweep prioritised so the readable expanse was banked first while the card was at its strongest. The regions holding the image content were revisited on later, patient passes, recovering sectors that had returned nothing to any ordinary read. The photographs were rebuilt from those recovered sectors, validated by rendering at full size rather than counted, and delivered ordered by their embedded capture dates.

The outcome

The card imaged under timeout control with the weak regions re-read, and the photographs rebuilt from recovered sectors and validated by rendering. 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, for anyone whose file list looks perfect and whose files do not open: the directory and the content are two separate reads from two different areas, so an intact listing over blank images means the small concentrated region still reads and the wider one does not; that is unreadable rather than corrupt, which is the better of the two because unreadable sectors often yield to patient repeated reads; the freezing is your software having no way to stop waiting; and both the camera and the computer failing identically proves the fault is the card. Stop copying — it costs you.

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.