ClipWhere does ClipLogger keep its own data, and what can I safely delete?
Data can exist beside media, in project documents, in ~/Library/Application Support/ClipLogger/, and in ~/Library/Caches/ClipLogger/vault-dbs/. Startup-disk and network-share catalogs use the Mac’s cache directory; local external volumes can carry <volume>/.cliplogger/index.db. Some derivatives can be regenerated, but catalog-only decisions, recovery records, and project state must not be treated as disposable caches. Quit the app and any CLI writer before making a consistent backup of database files and their companions. Use the app’s maintenance controls rather than deleting an index on the assumption that every value is already in a sidecar.
What exactly should I back up?
Back up camera media and all of its required companion files, .logger.json and enabled handoff outputs, project documents, and ClipLogger's relevant application data. Preserve both on-volume and app-folder projects. Keep recovery and catalog data when a portable write is pending or JSON output is off. Prove a restore on a small sample before relying on the archive.
Can two editors point at the same NAS folder at the same time?
Yes. Both editors can browse media on a shared NAS without importing another copy. Keep the storage mounted and coordinate changes to the same logging records. ClipLogger adds browsing, search and logging to that shared footage. Preserve project documents and enabled sidecars when moving work between machines; keep each Mac’s local catalog separate.
What happens when I delete clips inside ClipLogger?
Rejecting a clip is an editorial decision and is separate from recycling or permanently deleting a file. The recycle workflow identifies the affected media and reports the actual result. Inspect the list and scope before acting, and keep a verified backup before permanent removal. Nothing about an AI quality score should be treated as permission to erase an original.
Can I log one section of a long clip as its own clip?
Yes. Set in and out points and promote the range into a virtual clip, which you can browse, tag and analyze on its own without copying the camera original. Keep the parent file available. If you need a separate media file, export a subclip: stream-copy may start at a keyframe, while re-encoding gives a frame-accurate cut. Check source timing in your NLE export.
Is there a command-line tool?
The app repository includes headless tooling for specific scan, verification, and export workflows. Command availability depends on the build and distribution; consult the installed tool's help rather than copying internal developer commands into a production workflow. Do not run headless mutations against a library currently owned by another writer.
Do I need to keep the app open while a Rush job runs?
Keep it open through submission and the evidence upload; after that the batch runs server-side. Jobs still in flight when you quit are persisted, and on relaunch the app resumes polling them and ingests the finished results into the normal review flow. Cancelling from the Rush panel stops undispatched clips and refunds their credits.
What languages can transcription handle?
Language coverage depends on the selected transcription model. The English-only Whisper configuration and a multilingual model are different choices. Inspect Settings → AI for the installed model and supported setup path. Try speech in the actual language and acoustic conditions before trusting a long batch.
Can I run ClipLogger with AI switched off entirely?
You can browse and manually log without reasoning models. To review an AI-off configuration, inspect automatic text/speech analysis, reasoning activation, task routing, and any independently configured local engine. One cloud-named toggle does not necessarily disable an active local route. Stop existing work in Jobs and verify the resulting state before assuming all analysis is inactive.
How do I audit what the AI actually did on a clip?
Inspect the clip's provenance and available AI attempt history for provider, model, status, and the evidence associated with the result. Older records may not contain a complete route history. A missing record should remain unknown rather than being labeled local. Re-reasoning creates a new attempt; preserve the old evidence when comparing outcomes.
What does the Duplicates mode actually compare?
Choose the scan scope and available criteria, then review the proposed groups. Name, size, duration, metadata, similarity, and content-related checks serve different purposes. A candidate group is not an automatic delete instruction, and a similarity match is not file ownership. Review the proposed keeper and the original locations before sending anything through the recycle workflow.
Where do I see the app's background work?
The Jobs popover in the toolbar. Every long-running task is a job with progress and a stop control: subject scans, Prepare batches, duplicate scans, pre-caching, and Rush submissions each show there, and the Artifacts section controls background derivative work like deep pre-caching. Job status lines also surface inline near the feature that started them.
Where is the full keyboard shortcut reference?
Open the application's Keyboard Shortcuts window for the current app-wide bindings. Custom logging keys belong to the active palette and can differ by project. Scope matters: Command-Return confirms suggestions in one context and applies rename in another. Text input and the focused surface affect which command receives a key.
What does the app connect to on the network if I never sign in?
Potential connections include updates, requested model downloads, configured endpoints, and enabled connected-storage sharing. Signing out is not a master network-off switch. Review AI routing and Sharing & Privacy, then verify the intended workflow with the actual storage and endpoint configuration. Localhost and a server on your network are different destinations.
How do app updates work?
The application uses its update feed and updater to offer available builds. Check the release identity and requirements before installing on a production Mac, and preserve a backup of the relevant app data. A documentation preview is not evidence that a particular build is released, notarized, or approved for your workflow.
Can I import metadata from an existing Premiere or Resolve project?
There is no direct Premiere or Resolve project importer. ClipLogger can read supported XMP values, import roster or code-lookup CSV/TSV files, and recover fields from a known filename template. These serve different purposes; a lookup import does not recreate another application’s clip catalog.