ClipCatalog is a Windows application from Paulus Digital Solutions LLC for cataloging and searching video collections. It analyzes footage locally to identify visual content, transcribe speech and optionally group recurring faces. The commercial license includes a free trial with the same functions.
A catalog across drives
Users add folders from internal disks, external drives, SD cards or network storage. ClipCatalog scans the videos and records their paths, technical information and thumbnails in a local catalog. AI processing adds further search information without moving or renaming the original files.
The catalog remains searchable after an external drive disconnects. ClipCatalog recognizes volumes through stable identifiers and relative paths when they reconnect. Opening or copying an original still requires its drive. A folder sync collects new footage, and per-folder schedules can repeat scans while the application runs.
Finding scenes through descriptions
ClipCatalog analyzes frames and assigns AI tags for objects, scenes and activities. The Tags filter combines these labels with personal tags. All matching requires every chosen term; Any accepts a match to at least one.
Natural-language search accepts a description of the shot and compares its meaning with the visual index. Relaxed, Balanced and Strict settings change the breadth of matching. Users can sort results by semantic relevance and combine the description with other filters. AI tags identify general concepts, without identifying specific brands or product models.
Speech search and captions
ClipCatalog extracts each video’s audio and runs local speech recognition. It saves words with their timing information, then searches that transcript when users enter spoken words or phrases. Results highlight matching words inside each clip’s transcript. All and Any controls also apply to multiple speech terms.
Additional controls filter by transcription language or set minimum and maximum speech coverage. The latter measures how much of a clip contains speech. Users can export the transcript as plain text or SRT subtitles, or copy its text to the clipboard. Background noise and overlapping voices can reduce transcription accuracy.
Recurring faces
Face detection starts off, and users enable it in Settings after a confirmation. ClipCatalog detects faces in sample frames, compares their characteristics and groups repeated appearances into people. Selecting people adds a reusable person filter to the search. All requires every selected person, while Any accepts any of them.
The People view lets users correct groups and merge appearances of the same person. Brief appearances and small faces can escape detection. Users can disable the function and delete its face data from Settings.
Combining production filters
ClipCatalog combines content searches with file information and calculated categories. Its filters cover several different kinds of evidence:
- Recording dates, duration, resolution and frame rate narrow the technical properties.
- Orientation and audio controls separate horizontal or vertical footage and clips with or without sound.
- Camera filters read model information, while Drone checks camera and flight metadata rather than the picture.
- Places uses a video’s GPS position to identify its country, region and nearest town locally.
- Footage type estimates dialogue, voiceover and scenic shares from speech, faces and visual changes.
Users can save filter combinations as search presets. Footage type requires face analysis and AI tags, plus transcription for clips that contain audio.
Immersive footage
ClipCatalog recognizes native .insv, .360 and .osv files alongside equirectangular video. It extracts flat viewpoints and applies its AI indexing to those views. The 360 filter includes or excludes immersive footage. Turning off native-format processing keeps those files in the catalog without AI indexing.
Viewpoint sampling can miss action overhead or underfoot. ClipCatalog provides no panoramic viewer or reframing tool; those operations require another player or editor.
Previewing and handing off matches
The built-in player previews a matching video inside ClipCatalog. Checkboxes and range selection collect several results for one operation. Users can drag the selected original clips into a video editor or copy them to a chosen folder. These actions use the original files, so the corresponding storage must remain available.
Analysis hardware
ClipCatalog offers a primary AI device setting and separate overrides for tagging, face recognition and transcription. A benchmark tests the CPU and available GPUs, reports their processing speeds and recommends a device for each task. It shows proposed changes before users apply them.
Processing can pause and resume without discarding its progress. When GPU processing fails or becomes unavailable, ClipCatalog can continue that work on the CPU.
