NEW IN VERSION 8

Find Duplicate Photos by What They Look Like

Photo libraries are full of duplicates that no checksum will ever catch. Resize an image, convert it from JPEG to HEIC, re-save it from an editor, or let a phone strip its metadata, and the file changes at the byte level even though the picture is identical. Duplicate File Detective 8 solves this with visual comparison.

Visual comparison derives a compact perceptual key from the decoded pixel content of each image rather than from its raw bytes. That key is stable across metadata edits, format conversion, recompression, and resizing - so copies of the same photo group together even when their files share nothing in common. Photos that record their orientation in EXIF metadata are turned upright before the key is derived, so a copy that had that rotation applied to its pixels still matches the original. Because the key behaves like any other hash, it flows through the existing comparison engine, appears in the Hash column, and is carried into every export.

Visual comparison is deliberately conservative. It finds copies of the same picture, not pictures that merely resemble one another: cropped, retouched, watermarked, or genuinely re-oriented images are different pictures, and are not grouped. Images carrying too little visual detail to identify - solid colors, blank scans - are left out rather than swept together. As with any comparison method, review a duplicate group before removing anything.

Prefer to match on camera data instead? Duplicate File Detective 8 can also compare photos by image metadata - Date Taken (with sub-second precision, so burst frames stay separate), camera make, camera model, and pixel dimensions - individually or in any combination.

Side-by-Side Image Comparison

Finding candidate duplicate photos is only half the job - you still have to decide which copy to keep. The new image comparison panel presents every member of the selected duplicate group as a row of thumbnails, each captioned with the details that actually drive the decision: file name, pixel dimensions, file size, and Date Taken.

Thumbnails load in the background so the panel never blocks the user interface, tiles can be marked directly from the strip, protected files are clearly identified and cannot be marked by accident, and a slider adjusts tile size to suit your display. The panel docks anywhere in the main window and can be disabled entirely if you don't need it.

The Duplicate Folder Report

Users have long asked a question the duplicate file report couldn't answer directly: which folders are copies of one another? The new duplicate folder report answers it. Working from the results of your duplicate search, it reports every pair of folders that share duplicate content, along with how much of each folder is duplicated in the other, the space you'd recover by removing either side, and a confidence rating.

That confidence rating matters. Folder pairs are classified as complete, partial, or incomplete based on actual scan coverage, so a folder that was only partially scanned - or one affected by search filters or enumeration errors - can never masquerade as a complete match. Percentages are always calculated against the files that were present, meaning filtering can't inflate a folder's apparent duplication level. You can restrict the view to complete matches only, and mark either side of a folder pair with a single click.

Find Duplicate Documents by Their Properties

The same report saved twice by two different people is the same document - but it is not the same file. Office and PDF files embed authorship, revision counts, and save timestamps directly in the document, so two copies of identical content produce entirely different checksums.

Duplicate File Detective 8 adds document property matching, which compares documents by the properties that describe their content rather than their history. Available criteria include title, subject, page count, word count, character count, embedded creation date (which survives copy and save-as operations, unlike file system timestamps), and PDF document lineage. Choose any combination that suits the way your documents are managed.

Smarter File Name Matching

Two new file name comparison options target the near-duplicate names that show up in every real-world file system:

• Ignore copy and version decorations - strips the clutter Windows and applications add when files are copied, so Report.docx, Report - Copy.docx, Copy of Report.docx, and Report (2).docx all compare as the same name.
• Match words in any order - compares the set of words in a file name independent of their arrangement, so 2020 Beach Holiday.jpg matches Holiday - Beach 2020.jpg.

Both options work alongside the existing name normalization, digit-skipping, and partial-name comparison modes introduced in earlier releases.

A Redesigned Project Wizard

The new project wizard has been rebuilt around project templates presented as selectable cards, making it far clearer what each type of search actually does. In addition to the familiar quick, checksum, strong checksum, and duplicate song searches, version 8 adds dedicated duplicate photo and duplicate document project types. The photo template lets you choose its matching basis right on the card - visual comparison (the default) or image metadata.

The wizard's path selection step has also been reworked for easier, more predictable use.

Far More Powerful Results Filtering

Results filtering now understands ranges and comparison operators in addition to the wildcard text matching it has always supported. Filter expressions such as >100MB, <=2GB, 1MB..50MB, >=2024-01-01, or 3..10 can be applied to any size, date, or numeric column - including file size, size on disk, created, modified and accessed dates, group size, group file count, and music track, year, duration, and bitrate.

Size values accept units (KB, MB, GB, TB), dates accept both ISO format and your Windows locale format, and thousands separators are tolerated throughout. Anything that isn't a range expression continues to work exactly as before.

A Greatly Expanded Preview Panel

The preview panel received more attention in this release than in any before it. It now handles the file formats people actually have:

• Modern image formats including HEIC, AVIF, WebP, JPEG XR, and camera RAW (subject to installed Windows codecs).
• Image previews now honor embedded EXIF orientation, so photos display upright.
• Modern audio and video formats including MKV, FLAC, WebM, and AAC.
• Large audio and video files now play regardless of the image preview size limit.
• Zip archive contents are listed directly in the preview panel, including password-protected archives.
• SVG images now scale to fit the panel.
• Audio tag details are shown when media playback isn't available.
• Text file previews now show substantially more content.
• Binary files no longer render as garbled text.

Just as importantly, the preview and file property panes now load their data off the user interface thread . Selecting a large or slow-to-read file no longer stalls the main window.

SmartMark Enhancements

SmartMark is where large result sets turn into decisions, and version 8 makes it both safer and faster:

• New Mark All Files Outside Master Path(s) action - a single command that protects your master locations and marks everything else.
• SmartMark notices now report exactly how many files were marked and how many were skipped.
• Substantially improved SmartMark performance with very large result sets.
• SmartMark operations no longer alter files hidden by an active report view filter.
• Folder-based marking now matches target folder paths exactly.
• Clearer size-based SmartMark menu item labels.

Automation, Scheduling & Command Line

Unattended operation is now genuinely dependable. Command line runs return a process exit code, so scheduled tasks and batch files can finally detect failures. Runs always terminate when finished - the new /NoExit switch keeps the program open instead, when that's what you want. Unexpected errors during a command line run are written to the event log and reported through the exit code.

The task scheduler received matching attention: the task list identifies its own tasks more reliably, shows tasks that have multiple triggers or actions, reports failures from the Run button instead of ignoring them, and no longer flags other programs' scheduled tasks as orphaned. Monthly schedules set to a particular day of the week now run on the day that was actually selected.

Performance

The duplicate classification data store has been rebuilt on a typed schema, replacing the text-serialized format used by previous releases. The result is smaller records, faster sorting, and no per-file parsing overhead during the comparison phase - a measurable improvement on large scans, and the foundation on which the new image, document, and perceptual matching criteria are built.

• Zip compression now supports hardware acceleration.
• Preview and file property panes load data off the user interface thread.
• Excel exports automatically create additional worksheets as needed to capture very large datasets.
• The duplicate search progress dialog represents workloads far more accurately.
• Updated PDF rendering libraries.

Interface Refinements

The Comparison Options dialog has been fully redesigned to accommodate the new matching criteria while remaining approachable, and it now manages the relationships between options intelligently - selecting visual comparison, for example, automatically disables the byte-for-byte and size options that don't apply to it.

• Theme-related improvements across all application controls, including print and print preview in dark theme.
• Redesigned search path indicator badges.
• Cleaner marked and locked checkbox glyphs in the report view.
• Improved detail report group contrast in dark mode.
• Restored main duplicate search report font size option.
• Improved Hash Calculator dialog layout.
• New folder button in the folder browser dialog.
• Ribbon bar button for resetting docking pane positions.
• New version availability is now reported directly in the main window.
• Improved Duplicate File Manager layout and error handling.

Updated Documentation

Duplicate File Detective's help content has moved online and has been extensively updated and expanded. Context-sensitive help throughout the product links directly to the relevant topic, so documentation stays current between releases rather than aging inside an installed help file.

And Much More

Version 8 also includes a long list of smaller improvements and bug fixes - covering exports, marking behavior with active filters, project file compatibility, command line switch handling, and more.

• View the full product release notes for the complete version 8 change list, plus the version 7 history.