I’m trying to recover files from a failing hard drive with bad sectors, but the copy speed keeps dropping to 0 and the transfer never finishes. The drive has important data I really need to save, and I’m not sure if this is a software issue or a sign the drive is dying. Looking for advice on the safest way to recover the data without making things worse.
When I saw a CRC error the first time, I made it worse by poking at the drive for another hour. Bad move. If your PC freezes when you open a folder, or you get a “cyclic redundancy check” error, stop using the drive now. If it’s external, unplug it. If it’s your main system drive, shut the machine off. Every extra spin-up and every retry on damaged spots puts more strain on a drive that already looks shaky.
From what I’ve seen, this usually comes down to one of two things.
Soft bad sectors are the less ugly kind. The data and the error-check info no longer line up, often after a power loss, a forced reboot, or some random interruption. The surface might still be fine.
Hard bad sectors are worse. That points to physical damage, stuff like platter wear or a head going bad. No software fixes physical damage. The best tools only work around it by avoiding those areas.
One thing I would not do first is run CHKDSK if the files matter. People suggest it fast, and I get why, but it’s built to repair the file system so Windows can use the disk again. It is not built around saving every possible file. I’ve seen it clean up file references in a way that made later recovery harder. If the goal is getting your data back, the safer path is to clone the drive first, byte for byte, or make a full image. You take one careful pass at reading what still responds, then do the rest of the work on the copy.
If you want to try this yourself, recovery software makes more sense than repair tools. Disk Drill is one option I’d look at for a drive that keeps hanging. The useful bit here is its backup behavior on unstable media. It doesn’t sit there forever choking on one unreadable block. It skips rough spots, grabs the easy data first, then circles back and retries the bad areas with smaller reads. That matters. Less thrashing, less time wasted, better odds of pulling more off before the drive degrades further.
The workflow is pretty simple on paper.
Get a healthy target drive with enough free space. Make a sector-by-sector image of the failing disk. Scan the image, not the original, for recoverable files. Once the image is done, leave the bad drive powered off. Save recovered files onto a different third drive, not onto the original disk and not onto the drive holding the image. People skip this part and then wonder why things got messy.
There’s also a point where DIY stops making sense. If the drive clicks, grinds, beeps, won’t spin up, or vanishes from detection entirely, I’d assume a mechanical problem first. Software won’t save you there. It might push the drive closer to dead. If the files matter more than the cost, send it to a recovery lab like Gillware or Techchef. They have the tools and cleanroom setups for damaged hardware. The price usually hurts, often a few hundred dollars and sometimes a few thousand, but if the data is irreplaceable, that math changes fast.
Only after your files are recovered and checked should you bother trying to “fix” the original drive. At that stage, a full format or chkdsk /r might mark bad sectors so the system avoids them. I still wouldn’t trust a drive like this with anything important. Once a disk starts throwing warnings, I treat it like a bench drive at best. Not for photos, work docs, or anything you’d hate to lose.
What helped me after getting burned was setting up backups I didn’t have to think about much. The 3-2-1 rule is still the sane one. Keep 3 copies of your data. Use 2 different storage types. Keep 1 copy off-site or in cloud storage. It feels annoying right up until the day a drive starts dying. Then it feels cheap.
4 Likes
If copy speed drops to 0 and stays there, your drive is stalling on retries. Windows Explorer is bad at this. It waits, retries, hangs, then looks dead. Stop using normal copy.
I agree with @mikeappsreviewer on one part, don’t run CHKDSK first. I disagree a bit on one thing though. If the drive is still detected cleanly and not making nasty sounds, I’d check SMART before doing anything else. Use CrystalDiskInfo or smartctl. If you see lots of pending sectors, uncorrectable errors, or reallocated sectors climbing, treat the drive like it’s on borrowed time.
A few practical things:
-
Change how you connect it.
Use a direct SATA port if possible. USB bridges are flaky with bad sectors. Some reset the drive when reads hang, which kills long copies. -
Cool the drive.
Heat makes weak drives worse. A fan pointed at it helps more than people think. I’ve seen read stability improve after temps dropped from 48C to 35C. -
Copy smallest, most important files first.
Big video files hit more bad areas. Docs, photos, project files first. Triage matters. -
Use tools built for failure.
Disk Drill is a decent pick here because it handles unstable media better than Explorer. If you want to compare user feedback first, check Disk Drill user reviews and recovery experiences. If the drive keeps freezing during file copy, make an image and work from it. -
On Linux, ddrescue is still the thing I’d trust most for a raw clone.
Fast pass first. Retry later. Log file enabled. It’s ugly, but it works. -
Disable auto-mount and indexing if the OS keeps poking the drive.
Background reads waste precious time.
If the drive clicks, disappears, or takes minutes to identify in BIOS, stop. DIY gets riskier fast. At that point, each spin-up is a gamble. If it’s silent but slow, you still have a shot. Don’t keep re-trying the same folder for hours, taht usually ends badly.
0 MB/s during a copy usually means the drive firmware is stuck doing deep retries on a damaged area, not that Windows is ‘thinking hard.’ So yeah, Explorer is basically the worst tool for this job.
I mostly agree with @mikeappsreviewer and @hoshikuzu, but one thing I’d add is: stop browsing the drive in File Explorer over and over. Every folder preview, thumbnail, and metadata read can poke the same damaged spot again. People forget that and accidentally make recovery slower.
What I’d do instead:
- Turn off thumbnails in Explorer
- Do not open big folders repeatedly
- If it’s an external HDD, use a powered USB dock or different cable just to rule out power drop issues
- Check Event Viewer for disk errors like I/O, controller resets, CRC, etc.
- If the drive is readable at all, prioritize file types that matter most instead of whole-folder drag/drop
Also, not every hard drive bad sector issue explained here means the same thing. Sometimes the damaged area is tiny but sits right inside a file table or directory record, which is why the whole copy appears frozen. One ugly little spot can bottleneck everthing.
If you want a practical Windows route, Disk Drill makes more sense than normal copy because it can image unstable drives and then let you work from the safer copy. That part matters more than people realize. I would still save the recovered stuff to a totally separate healthy drive.
Small disagree with the ‘copy small files first’ advice though: if the filesystem metadata is damaged, hand-picking files can become a time sink fast. In that case, imaging first is still the cleaner move.
If the drive starts clicking, disappears randomly, or slows the whole PC to a crawl even at boot, stop DIY. That’s where ‘one more try’ turns into ‘well, crap.’
