My hard drive started making strange noises and now it has bad sectors, so I can’t access important files like photos, work documents, and backups. I’m looking for safe data recovery tips, software recommendations, or advice on whether I should stop using the drive and try professional hard drive recovery before the damage gets worse.
Hard drive with bad sectors. What I’d do first
I went through this once with an old 2 TB HDD full of photos and project files, and yeah, the first mistake is usually leaving it running too long. If your drive is throwing bad sector warnings, freezing Explorer, or dragging the whole PC down, stop using it now. No extra boots. No random scans. No copying stuff around to “see what still works.”
If it’s external, unplug it.
If it’s internal, leave it alone and use another machine for downloads and research.
What bad sectors usually mean
There are two buckets here.
Soft bad sectors:
These are logical errors. The data in a sector no longer matches what the error correction expects. Annoying, but sometimes recoverable with software.
Hard bad sectors:
This is physical damage on the disk surface or failing hardware. Age, shock, heat, worn heads, any of it. If you hear clicking, grinding, chirping, or beeping, I’d stop right there. DIY stuff tends to make it worse. At that point, a recovery lab is the lane.
If the drive still shows up in BIOS, Disk Management, or recovery software, and it’s “only” unstable, you still have a shot.
Don’t scan the failing drive first
This part matters more than people think.
A normal file recovery scan hammers the disk. On a healthy drive, fine. On a dying one, bad idea. I learned this the dumb way. A long scan pushed one of my old drives from “glitchy” to “not detected.”
What worked better was making a full byte-for-byte image first. Sector by sector. Raw copy. Then I unplugged the original drive and did the rest on the image file.
That gives you one big advantage. You only stress the bad drive once, or at least less.
Tool I usually see people use
If you want the simple route, I’d look at Disk Drill. The useful part here isn’t the usual file scan. It’s the byte-to-byte backup feature. On damaged drives, this matters because it tries to work around unreadable areas instead of faceplanting on the first bad block. Smaller block retries helped me on one failing USB drive. Slow, but it got farther than the other stuff I tried.
It also reads common file systems like NTFS and FAT32, so for normal Windows drives you’re usually covered.
What I’d do, step by step
-
Install the recovery software on a healthy computer.
Do not install anything onto the failing drive. -
Connect the bad drive and a second healthy drive with enough free space.
You need room for the full image. -
Use the Drive Backup or byte-to-byte image feature.
Let it finish. Don’t poke at the disk while it runs. -
Once the image is created, disconnect the failing drive.
At this point I’d put it aside and stop touching it. -
Mount or attach the image in the software.
Run scans against the image, not the original hardware.
If you recover the files, save them to a different healthy disk. Not back onto the problem drive. I know, obvious, but people do it.
Trying to reuse the drive after recovery
You can test it, sure. I wouldn’t trust it.
On Windows, chkdsk /r in PowerShell or Command Prompt will scan for bad sectors and mark them so the OS avoids them later. A full format, not quick format, also forces a sector check.
Still, once a drive starts collecting bad sectors, I treat it like a spare tire you already drove on too long. Maybe okay for junk data. Maybe scratch storage. Nothing important. Nothing you’d hate losing twice.
When software stops helping
If the drive is not detected at all, or the image is too damaged to pull files from, you’re down to lab recovery.
The rough price range I’ve seen is about $500 to $3,000, sometimes more if parts need swapping or platter work is involved. Ugly bill, but if the data matters, there isn’t much else left.
What I’d change after this mess
Set up 3-2-1 backups.
3 copies of your data.
2 different storage types.
1 copy off-site.
I used to think this was overkill. Then I heard head clicks at 1:30 a.m. and changed my tune fast.
3 Likes
Strange noises changes this for me. I’m a bit less optimistic than @mikeappsreviewer on the DIY side. If you hear clicking, scraping, or spin-up spin-down loops, skip long home recovery attempts. Those sounds often mean head or motor trouble, not a simple bad-sector issue.
What I’d do.
-
Check SMART from another system, read-only if possible.
Use CrystalDiskInfo on Windows. Look at Reallocated Sectors, Pending Sectors, Uncorrectable Sectors. If those are climbing, the drive is degrading fast. -
Avoid CHKDSK at the start.
I disagree with people who run repairs early. CHKDSK writes to the disk and moves data. On a weak drive, that can cost files you still had. Repair comes after recovery, not before it. -
Try a hardware-stable connection.
Use a direct SATA port if this is an internal HDD. If it is external, swap the USB cable and power adapter first. I’ve seen ‘bad sectors’ turn out to be a bad enclosure. Happens more than ppl think. -
Clone with a tool built for errors.
Disk Drill is fine for imaging and file recovery, especially if you want a simpler UI. If it keeps stalling, many recovery folks switch to ddrescue on Linux because it logs read errors and resumes cleanly. It’s ugly, but effective. -
Cool the drive normally.
Keep airflow on it. Do not freeze it. That old trick is internet junk. -
Prioritize file types if the clone fails.
Start with irreplaceable data. Photos, docs, client files. Skip media you downloaded elsewhere.
If the drive disappears from BIOS, or noise gets worse, stop. Lab time. Every extra power cycle hurts your odds.
For more real-world bad block recovery discussion, see practical bad sector data recovery tips from Reddit.
Strange noises is the part that pushes this out of normal “bad sectors” territory for me. I mostly agree with @mikeappsreviewer and @chasseurdetoiles on avoiding heavy scans first, but I’d add one thing people skip: check whether the failure is actually the bridge board or enclosure before assuming the platters are toast.
If it’s an external HDD, pull it from the enclosure only if you know what you’re doing and it’s not some weird hardware-encrypted model. A flaky USB-SATA board can cause read errors, disconnects, and fake-corruption symptoms. Seen it more than once. Same with weak power bricks.
Also, don’t keep rebooting to “see if it comes back.” That habit kills drives faster than ppl realize.
One more unpopular opinion: if the files are truly priceless, don’t do a marathon DIY session just because the drive still sorta spins. Pick a limit. Like one careful clone attempt, then stop. Every retry is gambling with the heads.
For software, Disk Drill makes sense if you want a cleaner way to create a disk image and then recover files from that image instead of beating up the original drive. That workflow is the key part. If you want a quick explainer, this Disk Drill data recovery walkthrough for damaged hard drives covers the basics in a more readable way than most forum posts.
Big thing: recover to a different drive, not the same one. And if it starts clicking louder, disappears from BIOS, or smells hot, stop messing with it. That’s not “maybe fixable,” that’s “lab bill incoming.”
I’m going to disagree a little with @mikeappsreviewer on one point: if the drive is making new mechanical noises, I would not spend much time even checking file system structure or browsing folders. Every directory click can force more seeks.
What I’d add to @chasseurdetoiles and @sonhadordobosque is this: watch the drive’s behavior, not just the bad sector count. If reads start fast and then drop to zero, or the drive disconnects under load, that often means instability beyond simple media damage. In that case, lower your ambition from “recover everything” to “get the irreplaceable first.”
A practical trick is to recover by file size and importance. Small docs, spreadsheets, project files, and family photos usually come off with fewer retries than giant video archives or VM images. Don’t start with the 200 GB backup file if you can avoid it.
If you do try software, Disk Drill is reasonable mainly because it lets you work from an image instead of poking the failing disk over and over.
Pros for Disk Drill:
- easier UI than command line tools
- can create a disk image first
- good for sorting recovered files by type
- less intimidating if you are stressed and in a hurry
Cons:
- not magic on serious mechanical failure
- can be slow on unstable drives
- deeper damage sometimes calls for lower-level tools or a lab
- licensing cost may sting if the drive is already half gone
One more thing people forget: check destination health too. Recovering onto a flaky USB drive or nearly full SSD is a dumb way to lose round two.
So my cutoff would be:
- detected, stable enough, no worsening noise = one imaging attempt
- clicking, vanishing, repeated spin resets = stop and send it out
And whatever happens, retire that HDD after this. Bad sectors plus noise is not a “maybe keep using it” situation.