What actually happens during data recovery.

Nothing useful can be quoted or attempted until we know what’s actually wrong, so every job starts with a free diagnostic. The drive is examined — powered carefully, listened to, and read at a low level where it’s safe to do so — to establish one thing above all: is this a physical failure or a logical one? A clicking drive with failed heads and a drive full of accidentally deleted files need completely different work, and cost completely different amounts. The diagnostic sorts your case into the right category, estimates how much is recoverable, and produces a fixed price in writing. Only when you approve that figure does anything else begin. If your drive is making noise or won’t be recognised, this is also the point where we can tell you whether the platters are likely intact — usually good news hiding behind an alarming symptom.

If the fault is physical, the drive has to be made readable before a single file can be copied — and that work happens in clean-air conditions, because an open drive exposed to ordinary room dust is a drive being slowly ruined. Failed read/write heads are replaced with matched parts from a donor drive of the same family. A seized motor is freed or transplanted. A burnt circuit board is repaired and the drive’s own unique calibration data carried across, because boards aren’t simply interchangeable. Corrupted firmware — the drive’s internal operating system, living in a reserved area you never normally see — is rewritten. None of this is about recovering files yet. It’s about getting the patient stable enough to read once. Logical failures skip this stage entirely; the hardware is healthy, so there’s nothing to repair.

This is the step people don’t expect, and it’s the most important one. Rather than reaching in and copying out your photos, we make a complete sector-by-sector clone of the entire drive first — an image — and then do all the actual recovery work on that copy. There’s a hard reason for it. A failing drive is dying a little more with every read, so it should be read as few times as possible: once, carefully, to make the image, working around the bad areas and coming back for the stubborn ones on gentler passes. After that, the fragile original goes back in its bag and is never touched again. Every reconstruction, every scan, every failed attempt and retry happens on the image, where mistakes cost nothing. If a lab offers to skip this and “just grab your files” off a failing drive, that’s the corner you don’t want cut.

With a good image in hand, the work becomes about the file system — the index that turns billions of raw sectors into folders and filenames you recognise. Sometimes it survives intact and the files come back exactly as you filed them. Often it’s damaged, and it’s rebuilt from the image, cross-referencing what’s left of the index against the data itself. For anything the file system can no longer point to, we carve files straight out of the raw data by recognising their signatures — the tell-tale byte patterns that mark the start of a JPEG, a Word document, a video. It’s slower and the filenames are usually lost, but the contents come back. For a RAID or NAS, this stage has an extra layer: each member disk is imaged, then the array’s original layout is reconstructed virtually before the file system is even addressed.

Recovered isn’t the same as intact, so before anything goes back the files are checked — opened, sampled, confirmed to be whole rather than a shell with the right name. We can show you a listing, or previews of the images, so you see what’s come back before you commit to anything. Then your data is written to fresh media — an external drive we post to you, or a secure download — and returned. On most single-device jobs, this is also the point where payment happens, because you’ve been on a no-recovery-no-fee footing the whole way through: if the files hadn’t come back, there’d be nothing to pay. Your original drive is returned too, or securely destroyed if you’d rather it was.

Two things fall out of understanding the sequence. First, why running recovery software on a physically failing drive is a genuine risk rather than a harmless first try: it forces the dying drive to work hard, over and over, across exactly the failing areas — the opposite of the read-it-once-and-image-it discipline that protects your data. Second, why a real recovery isn’t a flat, cheap number: the repair, the clean-air work, the careful imaging and the reconstruction are hours of specialist effort on equipment most people never see. If you’ve just lost data, the single most useful thing you can do is stop — power the drive down and leave it alone — and let a free diagnostic tell you which of these paths your drive is actually on. You’re welcome to bring it to the Princes Street lab, or post it to us from anywhere in Scotland.