My external hard drive with years of family photos suddenly became corrupted and won’t open normally. I’m worried that running the wrong repair tool or formatting it could make recovery impossible. What’s the safest way to recover photos from a corrupted drive without losing data?
If the drive still appears in Disk Management and the size looks right, I wouldn't assume it's dead yet. That often points more toward file system corruption than a totally failed drive, especially if it spins normally and isn't clicking, grinding, or dropping in and out.
I'd be careful with Windows offering 'Scan and fix', though. That button is tempting, but it's not the first thing I'd run on a drive with family photos on it. Repair tools can make changes to the file system, and if something goes wrong, recovery can get harder.
At this point, the goal should be recovery, not repair. Get the photos copied somewhere else first. Fixing the drive can wait.
One option that worked in a similar situation is Disk Drill. The useful part is that you can make a full image of the drive first, then scan that image instead of hammering the original disk over and over. If the drive is starting to fail, that matters, because you want to grab whatever is still readable while you can.
The preview feature is also helpful for photos. If the previews open normally, that gives you a decent idea that the files are actually recoverable and not just random filenames. The Windows version also lets you recover up to 100 MB for free, so you can at least test the scan before paying.
Once the important files are safely copied to another drive, then you can think about repairs. A quick format may be enough if it was just a corrupted file system. I still wouldn't trust that drive with anything important afterward, but it might be fine for temporary use.
A few other things can be worth trying, depending on what you're seeing:
- CHKDSK, but only after the data is recovered. It can repair file system issues, but it can also change things.
- TestDisk if Windows suddenly shows the partition as RAW or the partition table looks damaged.
- A different USB cable, enclosure, or computer if it's an external drive. Bad USB bridges and flaky cables can make a good drive look corrupted.
I'd stop with DIY software if the drive is acting physically unstable. Stuff like this is a bad sign:
- clicking noises
- disappearing and reconnecting
- wrong capacity shown in Disk Management
- lots of pending or reallocated sectors in SMART
- scans that slow down badly or keep freezing
That usually means the issue is hardware-related, not just a messy file system. In that case, repeated scans can make things worse. A proper recovery lab can image the drive with dedicated hardware, sometimes after temporarily fixing electronics or internal parts, then recover from the copy instead of stressing the original drive.
Labs are not cheap. Basic logical recovery can be around $300 to $600, while mechanical failures are often $800 to $2,000 or more, depending on the drive and the damage. The better labs often work on a no-data, no-fee basis, which helps a bit.
If it were my drive, I'd do it in this order:
- Check Disk Management and see if the capacity looks normal.
- Recover the photos to a different drive.
- Only after that, try CHKDSK, TestDisk, or formatting.
- Replace the drive if corruption comes back.
Drives are replaceable. Family photos usually aren't. Even if this one starts working again, I'd treat it as a warning and keep another copy somewhere else.
3 Likes
Do not save anything back to that drive. Even a “recovered” photo copied onto the same disk can overwrite another photo you have not recovered yet. I agree with @mikeappsreviewer that repair comes later, but I’d make the first purchase a new blank drive that is at least the same size as the bad one, then image the failing drive to that. If you are comfortable with Linux, ddrescue is built for making a best-effort clone from a flaky disk. If not, a tool like Disk Drill is easier because it can image first and scan the image. Either way, the bad drive should be read-only in practice: no CHKDSK, no format, no “initialize disk,” no copying recovered files back onto it.
Whether this is a spinning hard drive or an external SSD changes the risk a lot. If it is a normal mechanical drive, the main danger is making a weak drive read the same bad areas over and over. If it is an SSD or flash-based portable drive, the extra danger is that “repair” or cleanup operations can cause blocks to be discarded or remapped in ways that make recovery less predictable.
I agree with the general direction above: don’t “fix” the disk first. But I’d be even more boring about it. Before running any recovery scan, write down what Windows shows: correct capacity or not, RAW or NTFS/exFAT, drive letter or no drive letter, any weird noises, whether it disconnects. That tells you whether you’re dealing with a file system problem or a device that’s starting to fail. If the drive appears once and then vanishes, stop testing it like a stubborn printer. That’s when a lab is the safer choice.
If it stays connected and shows the correct size, make an image to a separate drive and work from that. Disk Drill is fine for this kind of user-friendly workflow because imaging first is the important part, not the brand name on the scan button. Just don’t let any program “repair,” “rebuild,” or “save recovered files” onto the bad disk. Recover to a completely different drive, then open a sample of the recovered photos at full size, not just thumbnails. Thumbnails can fool you into thinking the originals are okay when the actual JPEGs are partly damaged.
After the photos are copied and checked, then you can experiment with CHKDSK, TestDisk, formatting, or a new enclosure. At that point you’re only risking the old drive, not the only copy of the memories.
