If that's regularly happening then I would have to recommend that you install SMART software either from your disk drive manufacturer (for each hard disk in question, both the source and destination drive) and run it on each drive after reading the instructions. This can take hours to do the "long" test so you'll need to be patient. Unfortunately, many if not most USB hard disk interface chips do not support SMART so it may not work with external hard disks connected via USB. If this turns out to be the case with your external drive unit, then you may need to extract the hard disk from the external enclosure and put it directly in the PC (or any PC) to be able to do the SMART test.

The SMART test is the best hardware test you can perform on a hard disk that is suspected of potential corruption or failure, but it does take patience to wait for the full long test to complete.

Another option is to download and install the open source smartmontools software for Windows although it is less user friendly and more technical in nature: https://www.smartmontools.org/

Of course if you're using Linux, the software comes with the OS.

It's a good idea to have SMART software running 24/7 on every computer to monitor drives for pre-failure as the software can pop up a notification of pending drive failure or send you an email or other notification so you can replace the hard disk before it has a catastrophic failure. The sad thing.. is that even though our hard disks have had the hardware capability to inform us that they are on the downward spiral to failure for almost 20 years now, almost nobody knows about this feature other than seeing a notice during boot time and not knowing what it means, and Windows doesn't do anything useful with this information by default unless you know about SMART and install 3rd party software and configure it... so 20'ish years after the fact, most people still suffer disk failures and lose data which could have totally been prevented via early warnings of failure via SMART.

It's one of those sad things that drives me nuts. :) The second sad thing is that it doesn't work over lots if not most USB chipsets because USB vendors are cheapwads.

https://en.wikipedia.org/wiki/S.M.A.R.T.

Update: One thing I neglected to mention but should have in hindsight, is that due to all hard disks being very sensitive to vibration, movement and shock while they are running, drives in portable enclosures are much more susceptible to early death. The problem is that most people in general are unaware of this but the hardware is designed to hide it well too. All hard disks contain an area of disk that is reserved by the hardware for internal use for diagnostics and for hardware failure remapping.

What this does, is as bad sectors are discovered on the disk due to wear and tear or damage sustained to sectors caused by vibration or being jarred etc. the hard disk itself internally detects and remaps these bad sectors to good ones in the reserved area transparently to both the user and the operating system. Scanning for bad sectors will not discover this as it is at the hardware level. This is a good feature because it makes disks more reliable at the hardware level by making the errors invisible to the OS layer, but it means that a drive is sustaining damage which is being hidden and eventually it's 9 lives run out when the reserved area is full and it can no longer remap bad sectors anymore. Only then do newly formed bad sectors that occur end up being visible to the operating system and noticeable by OS supplied utilities that scan disks for bad sectors, however at this point the hard disk has sustained significant enough wear and tear over time and damage that it is usually on its last dying breaths or at best extremely unreliable.

So scanning with OS supplied utilities like CHKDSK, SCANDISK or equivalent wont tell the truth about the health of the actual hard disk hardware, just what the OS sees. Only SMART monitoring software can examine the hard disk's detailed health, and it does report on how many bad sectors the hard disk has internally remapped to the reserved area as well as many other factors that predict when disk death is likely near. So it is paramount to use SMART utilities for monitoring and reporting of disk health if you care about your data at all. :)

CHKDSK/SCANDISK can only tell you "yup, your filesystem is corrupted, want me to try and fix it?" or "Yup, your disk is already dead or losing data already" whereas SMART can tell you "your disk is still working and your data is in tact, but it has terminal cancer and you should be making funeral replacements right now and going to bars to look for a new mate". :)

Like how QR Codes are synonymous for using with phones for getting information on something, but no phones have the software installed so no one scans for the codes? (So QR codes are potentially going out of style even if they were heavily pushed on products for a year or two?)

I mean... QR codes don't work very well on machines lacking cameras... and none of my computers have cameras...

I never knew or realized that, but probably because I hate and have never owned a mobile phone, and have more or less ignored the entire QR code phenomenon although I know what they are. :) But yeah, that would be annoying. Not quite as annoying as owning a mobile phone, but close. :)

Whoops, I edited my larger post above to correct a few typos and fix up a few things and somehow it ended up being posted as a new quoted post, not sure how that happened. I re-edited the original and deleted the text that was originally in this post.

Uh.. I just meant this for normal backups\mirroring.
E.g: backup to a disk, then mirror to another one.

For example, SyncBackFree already checks file differences, but I think that it calculates hashes everythime, instead of storing&reusing the old ones.
Teracopy is good for verifying manual transfers, but it has the same problem.

.
No, I probably explained myself badly.
I just want to keep calculated hashes and reuse them for next transfers.
But I'd prefer a GUI program and I need mass checking.

sounds like an issue with the drive. i had a seagate external drive a few years ago that was doing this but thats because my kid picked up the drive and was messing with it while it was spinning (unbeknownst to me) so anyway, i just ended up getting a new drive.

Dah! second nature.

Thanks to everyone posting. There's a lot of good info here and I'll be sure to check out some of the programs suggested once the scandisk finishes. The HD is 3TBs that's hooked up to a Win7 computer that is a bit older so its going to take some time.

For the whole set (folder+subfolders or even the whole drive), or individual files?

If the former, dvdsig works pretty well for that purpose. Just a couple of clicks to scan and/or verify your files in the current directory. It doesn't need to be installed or anything, just copy it to the folder or drive (root) where you want to use it,

It is just that what to do if you want to add or change files in that set? To be sure, I guess you'd first have to verify the original set so that they are still ok, then do your stuff (add new files to the set or change existing ones), and then re-create the checksums for all the files.

Massive operations would be better.

Ideally:
- once a checksum of the original is calculated, it should be stored in the same path of the file.
- every new mirrored copy should be compared to that to save some time and to avoid losing the original correct result.
- Syncback should do all this automatically every time it detects changes or new files to move..

While storage drive errors are most common it could also be a faulty memory module that is corrupting your files.
I'd run Memory Diagnostics to check your RAM.

This is what was going on when I had pretty much this exact problem. While moving loads of stuff from drive to another, every now and then a file got unlucky and occupied a broken spot in memory during its transit through cache.

Memory errors are subtle and horrifying.