
Use darktable’s local-copy workflow to synchronize editing data back to the reconnected primary storage. A local copy is more than an arbitrary thumbnail cache: it lets the application continue using the source while the external drive is absent. Deleting cached material manually is not a substitute for checking that the current XMP reached the archive.
Establish which source darktable is using
The official local copies documentation explains that a local copy is used whenever present. It identifies those images with a white thumbnail marker and the tag darktable|local-copy. Use these indicators to inventory the travel set before deciding that the main drive already contains the current edit.
The manual locates local-copy operations in the lighttable’s Actions on selection module. Check the labels in the installed release. Keep the original drive, local copy and editing data intact while verifying the relationship, rather than sorting opaque cache filenames manually.
The RAW sidecar handoff guide explains why surviving camera pixels do not establish that a recent treatment survived too. This task concerns returning those instructions from darktable’s managed local copy to its original storage location.
Reconnect before retiring the local working set
Consider a hypothetical travel portrait edited on a laptop while its archive drive was disconnected. The photographer later sees the source file on that drive and assumes the trip edits are already there. The photograph’s existence proves source availability, not which processing instructions accompany it.
For a proposed return-to-archive check:
- Reconnect the original storage and confirm that the expected source location is available.
- Identify the intended local-copy images through their marker or tag.
- Preserve an approved reference and current editing data before any removal operation.
- Use the application’s documented local-copy synchronization workflow in Actions on selection.
- Verify the resulting sidecar and reopened edit from the primary-storage workflow before treating the travel set as retired.
The manual describes synchronization at startup when local copies and accessible original storage coexist. That helpful behavior is not a reason to skip verification after a failed mount, a changed path or interrupted travel workflow.
Compare the edit that mattered
Choose a representative frame with a distinctive crop and local correction. Compare those features against the retained reference and inspect the processing state. A matching basic exposure alone cannot establish that a detailed local repair also returned successfully.
Keep the approved-version reference identified throughout this check. Record the original location, travel editing period and which representative files were reopened. Do not rename the original path casually during diagnosis; the managed relationship depends on knowing which source is being returned.
Stop if synchronization cannot be verified
If primary storage remains unavailable or the expected editing data is missing, keep the local copy and investigate the application’s current status. Do not free disk space by manually deleting darktable’s cache directory while that copy may be the accessible working source.
The saved-state versus undo guide separates recoverable editing data from a past command list. This article reports no completed synchronization test. Accept the return only when the intended archived source can be opened with the required recent edit and the managed local-copy workflow can finish without losing that verified state.
Related tasks
- Why Does My RAW Photo Lose Its Edits on Another Computer?
- How Do I Keep the Approved AI Image or Video from Getting Lost Among Revisions?
- Why Can I Edit My Saved Layers but Not Undo Yesterday’s Steps?
Sources and scope
- darktable user manual: local: Local copies take precedence when present; Actions on selection synchronizes XMP to reconnected primary storage. Startup also synchronizes when originals are accessible.
Based on the current online darktable manual reviewed October 31, 2026. Match controls to the installed release. Examples and acceptance checks are proposed, not measured application or print tests.
Prepared by Imagild Editorial, the company editorial team responsible for these guides. AI assisted research and drafting. This is documentation-based guidance with proposed examples, not a claim of hands-on testing or an independent product review. Report a correction.
Cover: AI-generated conceptual editorial illustration; not a real application screenshot, observed output or before-and-after test.