What it is and why it matters

A hard drive stores data magnetically on spinning platters, read by heads that float a few nanometres above the surface. The computer addresses it as a long line of numbered sectors. Because the drive writes where it is told and keeps old data until it is overwritten, a hard drive is the most forgiving storage an examiner meets. Deleted files, old partitions and fragments of long-gone documents often survive for years.

That makes imaging the critical step. A bitstream image copies every sector, including the unallocated ones, through a write-blocker that guarantees nothing is written back. Hashes of the source and the image prove the copy is exact. From then on the original goes into storage and all analysis happens on the image.

Drives also hide things. The Host Protected Area and the Device Configuration Overlay can make a drive report fewer sectors than it has. Bad sectors can stall a copy. Failing heads, stuck spindles and burnt circuit boards need a recovery lab rather than an imaging station. Knowing when to stop and hand over is part of the skill.

Types of digital evidence

  • Allocated Files and File System Records

    The live contents of NTFS, FAT, exFAT, ext4 or other file systems, with their metadata and timestamps.

  • Deleted Files

    Content whose file system record was released but whose sectors have not been reused. Hard drives keep this for a long time.

  • Unallocated and Slack Space

    Free sectors and the unused tails of allocated clusters, where fragments of earlier files remain.

  • Deleted and Previous Partitions

    Remnants of partition tables and file systems from before a repartition or reinstall, often fully recoverable.

  • Hidden Areas

    Sectors behind a Host Protected Area or Device Configuration Overlay that the operating system never sees.

  • Drive Identity and SMART

    Model, serial number, firmware, power-on hours, power cycle count and reallocated sector counts that describe the drive's history and health.

  • Boot Records and Boot Code

    The MBR or GPT, volume boot records and bootloaders, which can show tampering, bootkits or a changed operating system.

Devices and data sources

  • 3.5-inch SATA drives from desktops and DVRs
  • 2.5-inch SATA drives from laptops and external enclosures
  • IDE and PATA drives from older computers still found in offices and homes
  • SAS drives from servers and storage arrays
  • External USB hard drives, including those with a built-in USB bridge that cannot be bypassed
  • Hard drives removed from NAS boxes and RAID sets
  • Drives with hardware encryption or vendor-specific enclosures
  • Damaged drives with head, spindle, firmware or circuit board faults

How an examination runs

The order matters. Each step protects the one after it.

  1. 01
    Document and Label

    Photograph the drive, record make, model, serial, capacity on the label and interface. Note any damage, noise or missing screws that suggest earlier opening.

  2. 02
    Read Identity and SMART

    Connect through a write-blocker and read the drive's identity data and SMART attributes before imaging. Compare the reported sector count with the label and the native maximum to spot an HPA or DCO.

  3. 03
    Choose Format and Tool

    Decide between raw and E01 images, compression and segment size, and pick an imaging tool or hardware duplicator suited to the drive's health.

  4. 04
    Image With a Write-Blocker

    Create the bitstream image, hash the source and the image and record both values. Verify by reading the image back.

  5. 05
    Handle Bad Sectors

    If reads fail, switch to a recovery-aware tool that skips bad areas first and retries later, keeping a map of what could not be read. Do not hammer a weak drive.

  6. 06
    Decide on Hidden Areas

    If an HPA or DCO exists, decide under authority whether to remove it temporarily to image the hidden sectors, document the native and current maximum addresses and restore the setting afterwards.

  7. 07
    Analyse the Image

    Read the partition table, mount or parse each file system, recover deleted files and partitions, carve unallocated space and build the timeline.

  8. 08
    Stop and Refer When Needed

    A drive that clicks, does not spin or is not detected goes to a clean-room recovery lab with its chain of custody intact rather than being powered on repeatedly.

Tools used in this discipline

Independent profiles: what each tool does, where it stops and what you need to know before using it.

Terminology

Sector and LBA
The sector is the smallest addressable unit of a drive, 512 bytes or 4096 bytes. Logical Block Addressing numbers them from zero to the drive's maximum.
Host Protected Area
A region at the end of the drive hidden by the ATA SET MAX ADDRESS command. The operating system sees a smaller drive. Used by manufacturers for recovery images and sometimes to hide data.
Device Configuration Overlay
An ATA feature that can change the drive's reported capacity and features. It sits below the HPA and can hide space from it as well.
SMART
Self-Monitoring, Analysis and Reporting Technology. Attributes such as power-on hours, power cycles and reallocated sectors that the drive maintains about itself.
Bad Sector
A sector the drive cannot read reliably. Drives remap them to spares and record the count. Clusters of bad sectors signal a drive that may fail during imaging.
Raw and E01 Images
Raw (dd) is a plain copy of the sectors. E01 is a container format that adds compression, segmenting and embedded hashes and case data.
MBR and GPT
The two partition table formats. The Master Boot Record is the older scheme with a 2 TB limit. The GUID Partition Table has a protective MBR, a primary and a backup header and CRC checks.
Clean Room
A dust-controlled environment where a drive can be opened to swap heads or platters. Opening a drive anywhere else destroys it.
SMR
Shingled Magnetic Recording, a layout that overlaps tracks for density and makes the drive manage rewrites internally. It slows imaging and can behave unexpectedly under heavy reads.

Challenges and limits

What makes it hard

  • Weak and failing drives can die during imaging, so the first attempt may be the only one
  • Hidden areas can be missed entirely if the examiner trusts the operating system's view of the drive size
  • Large drives take many hours to image and verify, which strains case timelines and storage
  • USB external drives with a built-in bridge may present a different sector size or hide ATA features, and the bridge itself may encrypt
  • Full disk encryption such as BitLocker turns a perfect image into an unreadable one without the key
  • Older interfaces such as IDE and SCSI need adapters that fewer labs keep
  • Sanitised drives give no deleted data at all, and the examiner must recognise a wiped drive rather than keep searching

Keeping the evidence intact

  • Image through a hardware write-blocker that has been tested on that interface and record the test
  • Record the native and current maximum sector counts before and after any HPA or DCO handling
  • Hash the source and image with SHA-256 and at least one other algorithm, and verify the image before analysis
  • Keep a read map for any drive with bad sectors so the report states exactly which sectors are missing
  • Store the original drive in anti-static packaging in sealed evidence storage once the image is verified

Standards and legal considerations

  • ISO/IEC 27037:2012

    The international guideline for seizure, acquisition and preservation that hard drive imaging procedures are built on.

    Read the guide
  • NIST SP 800-86

    Describes the collection and examination of data from storage media as part of a forensic process.

    Read the guide
  • SWGDE Best Practices

    Practical guidance on computer forensic acquisition, including the handling of hidden areas and damaged media.

    Read the guide
  • NIST SP 800-88 Rev. 2

    Media sanitisation guidance. An examiner uses it to recognise what a properly sanitised drive looks like and why nothing is recoverable from it.

    Read the guide

In practice

  • A Drive Smaller Than Its Label

    A 1 TB drive from an accountant's desktop reports 931 GB to the operating system but only 900 GB to the imaging tool. The native maximum shows a Host Protected Area of about 30 GB. Under authority the examiner images the hidden area and finds an older copy of the ledger database.

  • Reinstalled to Hide Activity

    A laptop was reinstalled the day before it was handed in. The image shows the new installation, but the examiner finds the previous NTFS volume's $MFT remnants in unallocated space and recovers file names, timestamps and several complete documents from the old system.

  • A Clicking Drive

    An external drive clicks and is not detected. Instead of retrying, the examiner documents the symptoms, seals the drive and sends it to a clean-room partner under custody. A head swap allows imaging, and the examiner proceeds from the image.

This page is educational material, not legal advice. Laws and standards change, commence on different dates and are interpreted differently across jurisdictions. Always read the official text linked on this page and take advice from a qualified professional for a real matter.

Learn it at DFRTI

Free pages explain. These courses teach it on real devices with a mentor.

Questions people ask

Why not just copy the files?

A file copy misses deleted files, unallocated space, slack, hidden areas and the metadata that dates everything. It also changes access times on the source. A bitstream image through a write-blocker captures everything and changes nothing.

What is the point of two hash algorithms?

Belt and braces. If one algorithm is ever questioned, the other still proves the image matches the source. Most labs record MD5 or SHA-1 alongside SHA-256 for this reason.

Can data be recovered after a format?

Usually, from a hard drive. A quick format rewrites the file system structures but leaves most data in place. A full format or a wipe that overwrites every sector is a different matter.

Should I open a failed drive to look inside?

No. Dust and fingerprints on a platter destroy data. Failed drives are opened only in a clean room by a recovery specialist, and only after the decision to do so is documented.

More in Devices and storage

Go deeper with DFRTI

Free pages explain the ground. Structured courses, lab work and a mentor take you from knowing about it to doing it on a case.