Data Corruption, Image Stream Corrupt or Byte-for-Byte Validation Failure Message

Image for Windows/DOS/Linux include the ability to reliably and accurately detect corruption of data from the source to the target and of the backup file itself (see Validation article).  Detecting corruption indicates that the software is working correctly and that something is not working properly on the system.  If this high level of detection wasn't built in to the software, detecting a corrupt bit could be hit or miss and you'd never know there was a problem. 

When data corruption is detected one of the following error messages will be displayed and/or logged:

ABORT: Data corruption was encountered (version 3.00 or later)

The image stream is corrupt (version 2.99-00 or earlier)

For validation error messages also see the following KB article: What is the Difference between "Validate", "Validate Byte-for-Byte", and "Validate Disc"?

The information below will attempt to help you find the cause of the corruption.

Byte-for-Byte Validation Failure when using VSS

Microsoft's Volume Shadow Copy service (VSS) may not protect everything that Image for Windows backs up. This will result in a byte-for-byte validation failure if any of that data has changed. The options "Exclude Files to Not Backup" and "Exclude Files to Not Snapshot" in the the Image for Windows VSS settings should be enabled (the default).  We have found the main culprit for this condition is that VSS is not protecting the $UsnJrnl.   To work around this you can either, a) Enable the option "Omit $UsnJrnl Data", b) Disable the option "Validate Byte-for-Byte", c) Enable the option "Ignore Validate B4B Errors", or d) use PHYLock instead of VSS.  In any case, we recommend at least running the memory check (explained below) on your system to ensure that bad memory is not the actual problem.

Byte-for-Byte Validation Failure when using a SSD

If the SSD (ATA) does not support Deterministic TRIM (e.g. DRAT, RZAT) then any data in allocated file system clusters that have been trimmed and only allocated (not used or written to) can change (with no real change) and cause the byte-for-byte validation to fail. Use a SSD that supports Deterministic TRIM to avoid this issue. Alternatively, use a NVMe drive, which requires consistent reads in trimmed areas as part of its specification.

Troubleshooting Corruption Issues

In general, corruption can be caused by overclocking, overheating, a hardware problem, or a BIOS or firmware bug or configuration issue.  The most common cause is bad memory and the second most common cause is a bad device port or cable.

To solve this issue:

First, ensure you are using the latest version of the software. An error such as "Not a valid image file (95)" typically means you created the image with a newer version, but trying to restore with a version that is too old. Additionally, some older versions may report the image is corrupt when trying to open a newer image.

Next, determine if the memory (or mainboard or CPU) is at fault. Run either the Microsoft Windows Memory Diagnostic Tool or memtest86+.  We find memtest86+ to be more thorough. Note that a memory tester may not catch all memory problems.

• memtest86+
  
  • Download memtest86+
  • Run all seven of the default tests and allow it to run overnight.
    
    
• Microsoft Windows Memory Diagnostic Tool (Windows Vista, Windows 7/8.x/10)
  
  • Open the Control Panel.
  • In the Search box, type in: memory
  • Click the Diagnose your computer's memory problems link.
  • Click the Restart now and check for problems option.
  • Run the extended tests overnight.
  • For more details, please refer to this Microsoft article.
    
    
• Microsoft Windows Memory Diagnostic Tool (Windows XP or earlier)
  
  • Download the Microsoft Windows Diagnostic Tool

If the memory diagnostics reports an error, replace the memory and repeat the memory test.  If you also get failures with new memory, and have ensured that the memory brand and type being used is compatible for the system, then there could be a problem with the CPU or mainboard.

If the memory diagnostic does not report any errors, check the following:

• Run chkdsk /f on both the source drive and the drive holding the backup image. Important: Only run chkdsk if the memory diagnostic did not report any errors. Running chkdsk on a system with bad memory can result in data corruption.
  
  Start a Command Prompt and then run chkdsk x: /f where x: is the drive to check.
  For example: chkdsk c: /f
  
  Note: If running Windows Vista or Windows 7/8.x/10, you must run chkdsk from an Administrator Command Prompt.
  
  
• If the problem is occurring when saving images to an optical drive, try burning at a slower speed and/or updating the drive's firmware, if possible.  Also, please refer to the article Message "Unable to read from file" when Performing Image Validation or Restore, if applicable.
  
  
• If you are using RAID, run a RAID consistency check or verify and fix (typically included in the RAID utility software).
  
  
• If you are overclocking, revert your settings to the defaults and determine if the problem goes away.
  
  
• Switch to higher quality internal and/or external cables.   
  
  
• Try a different port (internal or external) directly on the motherboard (not on the front of the case).
  
  
• Try different internal and/or external devices.
  
  
• If you have an IDE (PATA) drive that is set to master (and/or is the only device on the IDE cable) ensure that it is connected to the end of the IDE cable.  Also, PATA cables should be 80 conductor type and no more than 18 inches.
  
  
• If you are using a USB drive, please see the following KB article:
  Troubleshooting USB Drive Connection and Performance Issues
  
  
• Ensure that your system, including any applicable drives, are not running at temperatures that exceed the manufacturer specifications. If any temperatures are high, additional case fans may be needed.
  
  
• For Image for Windows, ensure the drivers for the storage controllers and devices being used are up to date.  Consider reinstalling the drivers even if the latest drivers are being used.
  
  
• Try a thorough surface test with a utility provided by the drive manufacturer, with SpinRite, or with a similar utility.
  
  
• Disable third party defragment utilities.   There are some third party defragment utilities that corrupt files larger than 2GiB on FAT32 partitions.  Defragment utilities can also cause corruption if the system RAM is bad.  The files are no good once they are corrupted by a defragment utility. 
  
  
• If using Image for DOS:
  
  
  • Attempt using the BIOS (direct) option instead of the BIOS option.  The BIOS (direct) option reverts to the BIOS option if it can't directly access the drive so BIOS issues may still apply.
    
    
  • If you are using a non-standard, third-party system BIOS or device firmware, switch to the latest final version of the manufacturer-supported BIOS or firmware, reset all BIOS settings to the defaults (referred to as a "clearing the CMOS" procedure by some mainboard manufacturers) and determine if the problem goes away.
    
    
  • If you have recently flashed to a different version of your system BIOS or the firmware of the controller to which your drive is connected (even if the version used is a manufacturer-supported version), try reverting to the BIOS or firmware version you were using before (or to the latest version) and determine if the problem goes away.
    
    

  • If you still have not narrowed down the problem, try adjusting the system BIOS parameters below, for the drive(s) with which you are experiencing the problem. These settings, if available, are typically found under "Integrated Peripherals" or in the "Drive Configuration" section.

    • Disable "Block mode"
    • Disable "32-bit access"
    • Switch to a different PIO mode
    • Switch to a lower UDMA mode, if using UDMA
      

If none of the above helps, you still may want to consider changing out the memory or even the CPU and motherboard.