Data Recovery Case File · Trust, Practice & Honest Limits · You Don't Need Us
Not a Recovery — a Copy, Done Properly
Her enquiry arrives here because two other trades sent her away. Over nine thousand photographs and around sixty short videos on a 32GB card, which she wants duplicated onto a second card, a USB stick and a 1TB external drive: "photography shops aren't willing to do this as it takes a long time and they are afraid if something gets corrupted along the way then they won't be able to retrieve it, so suggested I contact a data recovery firm." Nothing here is broken and this is not a recovery. But their nervousness is not unreasonable, and the answer to it is a method rather than a refusal.
| Media | 32GB SD card holding over 9,000 photographs and approximately 60 short video files — functional; duplication to multiple destinations required |
| Reported situation | Card functional with contents intact · duplication to a second card, a USB stick and a 1TB external drive requested · photography retailers declining the work on grounds of duration and risk · no fault present |
| Fault class | No fault — bulk duplication with verification required; source to be preserved as master until destinations are confirmed |
| Equipment used | Card read write-blocked so the source could not be altered · copied to each destination in turn · every destination verified by checksum against the source · sample images opened at full size on each · source retained as master until all destinations verified |
The decode: why they were nervous, and the method that answers it
What the shops were actually worried about, and it is legitimate: copying nine thousand files is not one operation but nine thousand, and a copy can fail silently. A file can transfer at the correct size with a bad block inside it, a transfer can drop partway through a folder and report success for what completed, and nothing in a progress bar distinguishes a good copy from a plausible one. Their real fear is the sequence that follows: somebody copies, assumes it worked, clears the card, and discovers months later that a third of the images do not open. They were declining the liability rather than the difficulty.
The method that removes the risk, and it is the same one film productions use: copy, then verify. Verification means computing a checksum — a short fingerprint — for every file on the source, computing the same on the destination after copying, and comparing them. Identical fingerprints mean identical files, byte for byte. Different ones name exactly which file went wrong so it can be copied again. Free software does this on every platform and it turns "I think it worked" into a result.
The order to work in: copy to the first destination and verify it. Then copy to the second and verify. Then the third. Keep the card as the master until every destination has verified, and do not reformat or reuse it before that. Nine thousand images at once is also worth doing in batches, so that a failure costs one folder rather than the whole job.
The point about her three destinations, which is worth knowing: they are not equally durable. Flash memory — the card and the stick — stores data as electrical charge that leaks slowly over years, faster in warmth, so a card in a drawer is quietly degrading rather than sitting safely. The external drive is the archive; the card and the stick are convenience copies. If only one is going to be checked periodically, check the drive.
The check nobody does: open a sample at full size on each destination — not a thumbnail, the actual image. Thumbnails are generated from previews and can display perfectly from a file that will not open.
Why this is worth writing down rather than quoting for: she can do this herself in an evening, and doing it herself means she will know how to do it again. That is a better outcome than paying somebody.
On the bench
The card was read write-blocked so the source could not be altered during copying — a precaution worth taking whenever the original is the only copy. Each destination was written in turn and verified by checksum against the source, so that a bad transfer named the specific file rather than passing silently. Sample images were opened at full size on each destination rather than viewed as thumbnails, and the source was retained as master until all destinations verified.
The outcome
The card duplicated to each destination with every copy verified by checksum, and the source retained as master until all confirmed. Free assessment, and no charge where no recovery is required. The decode, and the method: the shops were declining the liability rather than the work, and their fear is legitimate because a copy can fail silently — a file can transfer at the right size with bad content, and nothing in a progress bar tells you. The answer is verification. Copy, then compute a checksum on both sides and compare, one destination at a time, keeping the card as master until all of them pass. And treat the external drive as the archive, since flash loses charge over years.
Copying thousands of irreplaceable photographs
Verify each copy rather than trusting the progress bar — that's the whole thing, and it's why the shops were nervous. A copy can fail silently: a file can transfer at exactly the right size with a bad block inside it, and a transfer can drop partway through a folder while reporting success for the rest. The fix is a checksum, which is a short fingerprint computed for every file on both sides and compared; identical fingerprints mean identical files, and a mismatch names the specific file to redo. Free tools do this on every platform. Work one destination at a time, verify each before moving on, and keep the original card as your master until all of them pass. Open a few images at full size on each, not as thumbnails.
You can do this yourself — or call Newcastle Data Recovery on 0191 406 1051; source read write-blocked, every destination verified by checksum, samples opened at full size before the master is released.
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.