Every case file here is a real job off the Princes Street bench: what arrived, what had failed, how the recovery ran, and what went home. Eleven files are live below, rewritten properly from the bench archive — and more join them as the archive gives them up.
Bad sectors on both ReadyNAS drives, then a rebuild that overwrote metadata: virtual reassembly from images brought an architecture firm back 96%.
Seized spindle, implicated heads, a photographer’s deadline: donor parts and a careful image brought 98% back in 72 hours.
Dead controller, healthy flash: NAND chips lifted by micro-soldering, raw data decoded, 99% reassembled from a drive nothing could see.
Mirrors faithfully mirrored the encryption. Isolated, copied, virtually rebuilt — strain identified, 98% restored, attackers paid nothing.
512GB of unrepeatable photographs behind sudden corruption: forensic clone, rebuilt file system, 98%+ verified at full resolution.
Four striped drives, one head failure, no backup: donor heads and virtual reconstruction brought a financial firm back 97% whole in five days.
A second disk quit mid-rebuild on a Friday. Disks on the bench an hour after the call; trading again by Monday morning.
A researcher’s MacBook Pro, bad sectors under corrupt APFS, a deadline in days: imaged past the damage, 99% back in 48 hours.
Dead, silent and corroding: controlled cleaning and a micro-soldered board repair recovered 98% of a drowned LaCie Rugged.
A photographer’s entire archive — four undelivered weddings included — behind an electronic failure. Board-level repair; everything home.
A vintage case kept for its anatomy: donor head transplant, full image, 100% returned — on the data DVDs of the day.
No client is ever identifiable — devices, faults and outcomes are real; names and details that could point at a person are not published. Each file covers the same ground: the arrival story, the diagnosis in plain language, the bench work that followed, and the honest result including anything that couldn’t be saved. Between them they answer the question every visitor is really asking — ‘has this lab seen a failure like mine?’ — and for a bench with 25 years behind it, the answer is very nearly always yes. Beyond the files above, the process page shows how every one of them ran, and the reviews show how they ended.