Anti-Forensics Field Manual

This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong.
====== File Carving ======

===== Definition =====
File Carving is concerned with data that is deleted, partially recoverable, unrecoverable, or left behind in storage areas that no longer belong to an active file.
In a forensic report, file carving should be treated as a source of evidence and uncertainty, not as a shortcut to intent. The useful question is what the record supports, what it does not support, and what another source says when asked the same question.

===== Background =====
Deleted data is where forensic hope goes to negotiate with storage reality. Sometimes the file is recoverable. Sometimes the SSD has already swept the floor. Sometimes the carved fragment is just a few bytes wearing a convincing hat.
A reviewer should be able to follow the claim from source evidence to cautious conclusion without spelunking through unsupported confidence. A fragment can be useful, but it should not be promoted to witness of the year.

===== Technical Description =====
The technical work compares allocation state, remnants, metadata, carving output, recovery limits, storage media, and the timing of collection.
A careful review starts with the source format and collection method before it argues about meaning.

===== Forensic Relevance =====
  * **Timeline reconstruction:** File Carving can help place activity in sequence when the time source and collection conditions are understood.
  * **Recovery analysis:** It can explain why data was recovered, missed, corrupted, or only partially reconstructed.
  * **User activity review:** It may support account or profile context, but only when independent artifacts point the same direction.
  * **Contradiction testing:** It is useful for finding places where logs, metadata, storage behavior, or accounts disagree.
  * **Reporting decisions:** It helps decide how narrow the finding should be when file carving looks interesting but not conclusive.

===== Evidence Sources =====
^ Evidence source ^ What it may show ^ Reliability limits ^ What it cannot prove alone ^
| Unallocated space | May contain remnants of deleted files, old file fragments, or carving candidates. | Can contain stale, partial, or unrelated bytes from many time periods. | Does not prove when or why the data was deleted. |
| File-system records | May show deletion, rename, parent path, or sequence information. | Can be reused, truncated, or overwritten. | Do not prove deliberate destruction. |
| Recovery output | Shows what tools could reconstruct under examined conditions. | Depends on signatures, fragmentation, media behavior, and tool limits. | Does not prove that unrecovered data never existed. |
| Storage-device behavior | May explain why recovery failed, especially with SSD TRIM and wear leveling. | Often sits below the file-system view. | Does not establish user intent. |

===== Interpretation Limits =====
The most common error is treating file carving as proof of motive. It may support a technical event, a sequence, or a contradiction, but motive needs stronger ground.
Another bad leap is treating absence as intent. Missing data can come from retention, configuration, collection scope, sync behavior, overwriting, media behavior, or ordinary use.
A tool result should be described as a parsed or recovered record, not as the tool's opinion about what happened. Tools surface evidence; they do not understand it.
Failed recovery is a result, not a confession with a dramatic signature.

===== Common Misinterpretations =====
  * Ignoring benign explanations because the suspicious explanation is easier to write.
  * Treating recovery failure as proof that evidence was destroyed.
  * Treating one carved fragment like a confession because it fits the preferred story.
  * Ignoring SSD TRIM, overwriting, fragmentation, and collection timing.
  * Reporting a recovered file without explaining carving limits and validation.

===== Example Scenario =====
An examiner finds unallocated space related to file carving during a triage review.
The finding may support activity in the relevant time window, but it does not prove who caused it or why.
The report should state the observed record, the collection limits, and the alternative explanations that were tested.

===== Analysis Workflow =====
  - Define the question before opening another parser: what should File Carving help answer?
  - Preserve the source evidence and document how the unallocated space was collected.
  - Record tool versions, input paths, output paths, time settings, and errors.
  - Identify observed facts before writing any interpretation.
  - Normalize time sources and document timezone, clock drift, and collection-time effects.
  - Compare at least two independent artifact families before raising confidence.
  - Consider benign explanations, automated behavior, retention, sync, and storage-device behavior.
  - Write conclusions proportionally: observed fact first, inference second, uncertainty always visible.

===== Reporting Guidance =====
Reporting on file carving should be precise enough that another analyst can retrace the claim without inheriting the original examiner's confidence.
Avoid wording that converts possibility into intent. The report should not say a user deliberately deleted, hid, wiped, or tampered with evidence unless the evidence actually supports that conclusion.
Report-ready wording:
  * The available unallocated space is consistent with activity related to file carving, but it does not independently establish motive or user intent.
  * Recovery was limited under the examined conditions; additional corroboration would be required before concluding deliberate destruction.
  * The finding should be read with the collection scope, time-source limits, and alternate explanations described in this report.

===== Confidence and Reliability =====
^ Confidence level ^ What it looks like for this topic ^ How to report it ^
| Low | Unallocated space exists, but collection scope, time source, or surrounding context is limited. | State the observation and keep interpretation narrow. |
| Moderate | Unallocated space aligns with File-system records, but attribution or intent remains unresolved. | Say the artifacts support the finding, not that they prove it. |
| High | Multiple independent sources agree on sequence, source system, account context, and collection conditions. | Use stronger language, but still separate observed facts from inference. |

===== Related Concepts =====
  * [[wiki:digital_forensics:file_systems:carving_false_positives|Carving False Positives]] - Helps keep the finding grounded in a wider evidence pattern.
  * [[wiki:digital_forensics:file_systems:data_remanence|Data Remanence]] - Compare this concept when testing sequence and corroboration.
  * [[wiki:digital_forensics:file_systems:deleted_files|Deleted Files]] - Helps keep the finding grounded in a wider evidence pattern.
  * [[wiki:digital_forensics:file_systems:overwriting|Overwriting]] - Helps keep the finding grounded in a wider evidence pattern.
  * [[wiki:digital_forensics:file_systems:recycle_bin_artifacts|Recycle Bin Artifacts]] - Helps keep the finding grounded in a wider evidence pattern.
  * [[wiki:digital_forensics:file_systems:slack_space|Slack Space]] - Useful when checking whether another artifact supports the same interpretation.

===== Tools =====
  * [[https://www.autopsy.com/|Autopsy]] - Case review, file-system browsing, timelines, term search, and recovered artifact review.
  * [[https://www.sleuthkit.org/sleuthkit/|The Sleuth Kit]] - Command-line file-system analysis and evidence inspection.
  * [[https://www.cgsecurity.org/|PhotoRec and TestDisk]] - File recovery and partition recovery utilities.
  * [[https://foremost.sourceforge.net/|Foremost]] - File carving by headers, footers, and simple signatures.
  * [[https://github.com/sleuthkit/scalpel|Scalpel]] - Configurable file carving utility.
  * [[https://github.com/simsong/bulk_extractor|bulk_extractor]] - Feature extraction from disk images without relying on file-system allocation.
  * [[https://www.exterro.com/ftk-product-downloads|FTK Imager]] - Forensic imaging, preview, and export workflows.
  * [[https://guymager.sourceforge.io/|Guymager]] - Linux forensic imaging with hashing and acquisition logs.
  * [[https://sourceforge.net/projects/dc3dd/|dc3dd]] - Disk imaging and hashing utility.
  * [[https://www.gnu.org/software/ddrescue/|GNU ddrescue]] - Imaging damaged media while preserving recovery logs.
  * [[wiki:tools:start|Forensic Tools]] - Full tool directory for the wiki.

===== Limitations of Tools =====
Tools parse, surface, and organize evidence. They do not create conclusions.
Parser output can be affected by version differences, unsupported formats, corrupted records, timezone handling, partial collection, and storage behavior outside the tool's view.
When a tool produces a strong-looking result, validate it against another tool or source where the stakes justify it. A parser can recover a fragment; it cannot tell you whether the fragment deserves a paragraph in the report.

===== References and Further Reading =====
  * [[https://learn.microsoft.com/en-us/windows-server/storage/file-server/ntfs-overview|Microsoft NTFS overview]] - NTFS technical reference.
  * [[https://www.sleuthkit.org/sleuthkit/docs.php|The Sleuth Kit documentation]] - File-system analysis documentation.
  * [[https://www.nist.gov/itl/ssd/software-quality-group/computer-forensics-tool-testing-program-cftt|NIST CFTT]] - Tool testing context.
  * [[https://csrc.nist.gov/publications/detail/sp/800-86/final|NIST SP 800-86]] - Guide to integrating forensic techniques into incident response.
  * [[https://www.forensicfocus.com/articles/what-is-digital-forensics/|Forensic Focus: What is digital forensics?]] - Practitioner-facing overview.

===== See Also =====
  * [[wiki:digital_forensics:file_systems:start|File Systems]]
  * [[wiki:digital_forensics:file_systems:carving_false_positives|Carving False Positives]]
  * [[wiki:digital_forensics:file_systems:data_remanence|Data Remanence]]
  * [[wiki:digital_forensics:file_systems:deleted_files|Deleted Files]]
  * [[wiki:digital_forensics:file_systems:overwriting|Overwriting]]
  * [[wiki:digital_forensics:file_systems:recycle_bin_artifacts|Recycle Bin Artifacts]]
  * [[wiki:digital_forensics:file_systems:slack_space|Slack Space]]
  * [[wiki:digital_forensics:file_systems:ssd_trim|SSD TRIM]]
  * [[wiki:digital_forensics:file_systems:trash_folders|Trash Folders]]
  * [[wiki:digital_forensics:file_systems:unallocated_space|Unallocated Space]]
  * [[wiki:digital_forensics:file_systems:file_allocation_table|File Allocation Table]]
  * [[wiki:tools:start|Forensic Tools]]
  * [[wiki:resources:start|Project Resources]]

===== Reader Takeaway =====
File Carving is useful when it helps explain what the evidence can support and where the limits begin. Treat file carving as one part of a corroborated record, not a shortcut to intent. Failed recovery is a result, not a confession with a dramatic signature.

===== Use Notes =====
This article is for defensive education and technical reference. It should not be treated as legal, forensic, investigative, compliance, or operational advice without qualified professional judgment.