A cyclic redundancy check (CRC) error means that your computer has read data from the drive and it does not match the checksum stored with it. This means either there is an issue with the data itself or with the disk beneath it. In some cases this may be a one-off; often it indicates bad sectors on a drive that is starting to fail. Below we describe what the error means, what you should not do about it and how the data is recovered.
A cyclic redundancy check (CRC) is a maths check that data has not changed since it was originally written. When it fails, the bytes coming off the disk are wrong, generally because of bad sectors, a failing drive or a flaky cable. Once the CRC fails, copying stops, and forcing it to continue can make a failing drive worse.
chkdsk /f forces a failing drive to re-read and remap the bad areas of the disk. This places additional stress on a disk that is already struggling and can turn a recoverable drive into an unrecoverable one. If the data matters, image the drive first.When data is written to a disk, a small check value known as the cyclic redundancy check is also recorded alongside it. Each time that data is retrieved, the drive recalculates the check and verifies that it matches. A “Data error (cyclic redundancy check)” message simply means the two did not match: the bytes retrieved from the disk do not correspond with those that were originally written. This message appears on hard drives, external drives, USB sticks, SD cards and optical media alike. Although it does not necessarily mean the whole drive is lost, it is a clear sign that something physical is wrong, and it should be approached with care rather than repeated retries.
Bad sectors are the most common cause: areas of the disk that the drive can no longer read reliably, which multiply as a drive ages or begins to fail. A loose or failing cable can also cause this problem, and on an external drive the USB bridge inside the case is as likely a culprit as the disk, so try a different cable and a different port first. Less often, the fault is file-system corruption from improperly ejecting the drive or a power cut while writing. Looking at the pattern tells you a lot: a single CRC error that goes away is usually minor, but one that keeps returning or spreads to other files points to a disk that is deteriorating. On NTFS, that same bad-sector damage often surfaces as the “file record segment is unreadable” error during a chkdsk scan.
Windows will often suggest running chkdsk /f /r, and on a healthy drive with minor corruption this is a reasonable fix. On a failing drive, though, it is the wrong move. chkdsk forces the drive to re-read every sector and remap the bad ones, which hammers a disk that is already on the edge, and in the process it can overwrite the very data you want to keep. If you do not know whether the drive is sound, do not run chkdsk. Save whatever you can to another device, but the moment copying throws CRC errors or slows to a crawl, stop immediately and treat the drive as failing.
The safe route is the complete opposite of forcing it. The drive is imaged read-only, retrying and skipping the bad sectors so that the good data comes off without stressing the disk further, and the files are then rebuilt from this image instead of directly from the failing drive. A single hard drive is from £300 +VAT; an external drive is sent whole, because on a WD the encryption key resides on the bridge board and the bare disk is no use on its own; a memory card or USB stick is from £250 +VAT. Each job begins with a free diagnostic and a fixed written quote.
A cyclic redundancy check (CRC) is a maths check that is saved along with your data to confirm that it hasn’t changed. A CRC error means the data read back from the disk did not match that check, so the bytes are corrupt at the physical level. Most often a CRC error is due to bad sectors on a drive that is beginning to fail, but a loose cable or file-system corruption can also trigger it.
Only once you’re certain that your drive is healthy. Running chkdsk /f /r on a failing drive forces it to re-read and remap every bad sector, which stresses a disk that is already struggling and can overwrite recoverable data. Do not run chkdsk if the files matter and you do not know whether your drive is sound; image the drive first.
Usually, yes. The drive is imaged read-only, retrying and skipping the bad sectors so that the good data can be retrieved without stressing the disk further. Once the imaging is complete, the files are rebuilt from that image. A single drive costs from £300 +VAT after a free diagnostic.
Often it is the same bad sectors, but on an external drive it can also be caused by the USB bridge within the enclosure or by a failing cable rather than the disk itself. First, try a different cable and a different port. If the issue persists, send the whole unit: on a WD drive the encryption key lives on the bridge board, so the bare disk is no use on its own.
Not always, but it does serve as a warning sign. A single CRC error could simply be corruption or a problem with a cable, while one that keeps returning or spreading usually indicates bad sectors on a failing disk. If the drive is also clicking or slow, treat it as failing and recover the data before doing anything else.
If the error keeps returning or the drive is slow or clicking, do not keep running repairs over it. Send us your drive and we will image it read-only past the bad sectors, with a free diagnostic and a fixed quote before any work.