From Photo Mechanic to Lightroom, What Happens to Your Selections?

Why Photographers Love Their Tools · An editorial collection from Zippi Technologies. We develop Lightbox Pro; these are educational explanations, not independent product rankings or measured competitor tests.

First decide which Lightroom you mean

This procedure concerns desktop Lightroom Classic, not cloud-oriented Lightroom. It is based on the current official references linked beside the relevant steps. It is a documentation-verified procedure to rehearse on copies, not a claim that every camera format and installed version pair was run in our laboratory.

Photo Mechanic can provide a focused arrival and selection environment while Lightroom Classic handles catalog-based development. The boundary is not inherently unreliable. Problems arise when the photographer expects a decision to transfer through a channel that was never configured, or when two applications both have different current records.

Separate three things before beginning: the original file, the selection information and the preview used to view it. A correct rating can accompany an old-looking preview. A fresh-looking preview can accompany the wrong source. Neither visual appearance alone nor a count of selected thumbnails answers every transfer question.

The existing metadata compatibility chapter provides a general acceptance framework. This guide supplies a narrower operator procedure for one named handoff, including the important difference between a first import and rereading records for photographs already in a Classic catalog.

Build a six-file acceptance packet

Use copies of six photographs from formats you actually shoot. Name the packet and record the exact Photo Mechanic and Lightroom Classic versions, operating system, file extensions and whether any photograph is already cataloged. Retain an untouched copy of the packet before changing metadata.

Scroll within the table to read all columns.

The ratings are examples, not a recommended universal culling policy. Write the expected values in a small private worksheet. The worksheet is the reference against which both programs are checked; it should not be generated afterward from whichever screen happened to look correct.

Include one deliberately cleared field if your real process depends on clearing values. An empty field may mean “intentionally removed” or “never written.” The acceptance packet makes that difference visible before it affects a whole job.

Set the writer before making the selection

Review Photo Mechanic’s IPTC/XMP preferences before assigning the test decisions. Camera Bits documents RAW-sidecar defaults and an option controlling whether TIFF-based RAW originals may be modified. For a sidecar-based rehearsal, keep RAW modification disabled and confirm that the intended XMP sidecars are created or updated for the supported RAW types.

Use the relevant Camera Bits compatibility snapshot guidance as a starting configuration, then verify it in the installed version. Do not treat a snapshot name as proof that every field, proprietary flag or future software version interoperates automatically.

Choose a single writer for the handoff stage. While Photo Mechanic is establishing the selection, avoid making competing metadata edits to the same packet in Classic. The workflow is easier to verify when there is a known source of the intended state.

Check the resulting file set in the working folder. The RAW source and its corresponding sidecar need to remain together under the expected base name. A copy operation that transfers only the RAW file may omit the record the downstream reader needs. For embedded-metadata formats, the storage route differs; do not manufacture a sidecar requirement for every extension.

Match label text, not just the swatch

Lightroom Classic color labels use text definitions within a label set. Match the intended text between applications rather than assuming that a red-looking swatch in one means the same thing in another. Case, spelling and custom meanings belong in the packet’s record.

For the example, the team uses “Review” for an unresolved photographic question and “Approved” for a later production state. Keep those meanings explicit. If a local color assignment is changed, check the text that is actually exchanged before interpreting the receiving color.

Do not use Photo Mechanic tags and Lightroom pick/reject flags as the sole acceptance channel unless their exact route has been independently verified for the installed versions and file types. This guide’s baseline depends on documented stars and label text plus explicit selected membership. It makes no blanket claim that all proprietary selection signals translate.

A mismatch in color display may be a label-set interpretation problem rather than a lost selection. Inspect the underlying text and the configured set before rewriting the files. Rewriting first can destroy useful evidence about where the original discrepancy arose.

First import: bring in the existing source location

Finish the packet’s Photo Mechanic changes and allow its writes to complete. Confirm the expected files and sidecars are present. Then use Classic’s folder-import workflow. Where the intention is to catalog files already in their correct location, choose the supported Add route rather than creating an unplanned second working tree.

Review the import’s destination behavior and any applied metadata preset before proceeding. A preset that deliberately replaces a field can make a successful read look like a failed transfer. During the small rehearsal, keep unrelated automatic changes out of the test so the result has a clear explanation.

After import, select the packet’s exact folder or an explicitly bounded group. Inspect A, B and C individually, then compare the expected rating groups. Confirm D by its full location, not its basename alone. For E, determine which member of the RAW-plus-JPEG relationship you are actually viewing.

Record both the expected and observed state before making corrections. If a value differs, retain the packet and settings as diagnostic evidence. The purpose is to identify the transfer route, not to make the screen look right through an unrecorded manual adjustment.

Already cataloged: reading metadata is a deliberate overwrite

A photograph already present in Classic may have catalog state that predates an external change. Reimporting is not the right universal answer. Adobe documents Read Metadata From File and conflict resolution, including the fact that applying disk metadata can overwrite catalog settings. Treat the direction of that action as consequential.

On copied test work, preserve the catalog and associated file/sidecar state first. Identify exactly which record is intended to win. If Classic contains important developments or metadata not represented by the disk record, resolve that difference before reading externally changed settings across a large selection.

Use the documented reading command only on the bounded packet after reviewing the consequences. Recheck the existing development state as well as stars and labels. Do not promise that a general reread imports only a rating while leaving every other property untouched.

The safe operational pattern is: freeze the competing records, choose the intended authority, perform the smallest relevant action and inspect the result. If the correct authority cannot be determined, stop the handoff for those files and keep the conflict visible. A confident-looking selection created by overwriting the wrong record is not a successful transfer.

Automatic XMP writing is a different direction

Classic’s automatic XMP-writing option concerns writing its metadata outward. It is not an instruction for Photo Mechanic’s later changes to become authoritative in the catalog. Enabling outward writing as a generic “sync fix” can add a competing writer to the very boundary being investigated.

Decide whether the real workflow is one-way intake or a controlled round trip. A one-way intake can finish selection in Photo Mechanic, verify the read in Classic and then let Classic own subsequent work. A round trip needs a more explicit policy for which properties each program may change and when.

Keep the policy small enough to understand. For example, captions might remain managed in one program while photographic development remains in Classic, but that arrangement still needs a verified field-level exchange. A shared file location does not automatically create a conflict resolver.

Before adopting a round trip, modify a noncritical test rating in each direction and inspect every expected field after each action. Then deliberately create a disagreement and rehearse its resolution. The goal is predictable behavior under a conflict, not only a successful happy-path demonstration.

Reconcile identities as well as counts

Suppose the example job has 120 transferred candidates. A receiving count of 120 is necessary but insufficient: one missing intended frame and one accidental extra still produce 120. Compare the actual identifiers or source paths for the selected membership.

For the six-file packet, verify each item against the worksheet. For a larger assignment, preserve a list of selected identities at the handoff and investigate missing or unexpected members. Keep camera-folder context where basenames repeat. Do not reduce identity to the last four digits of a filename.

A hash can help confirm that a copied original is the expected file, but it does not by itself encode a later rating or make two differently packaged representations identical. Use each piece of evidence for the question it can answer.

When the preview looks different but the selection is right

Camera Bits explains that external RAW edits may not update the embedded preview displayed by Photo Mechanic. A different rendering after the handoff does not necessarily mean the wrong exposure was selected or the rating failed to transfer.

Confirm the identity and metadata first, then investigate appearance. Which program rendered the current view? Is it an embedded camera preview, a generated preview or the current developed interpretation? Has the intended editor actually saved a new rendition that the viewing application can display?

Do not judge transfer integrity solely by color, crop or brightness. Those can differ because proprietary development instructions are interpreted by their owning processor. Conversely, a very similar-looking photograph can still be a neighboring exposure. Identity and appearance deserve independent checks.

For detail-critical selections, inspect enough source detail in the appropriate application. A small thumbnail cannot certify eyelashes, fabric or motion. The sharpness guide addresses that photographic decision without confusing it with metadata transport.

A bounded troubleshooting order

If all ratings appear absent, confirm the exact folder, file type and writer configuration before editing anything. Look for missing sidecars, a different file copy or a catalog view filtered to another group. Then inspect a known positive from the packet.

If stars arrive but colors do not, examine label text and the receiving label set. If only previously cataloged files are stale, investigate the reading/conflict path rather than repeating a successful first-import experiment. If the values change again later, look for an unintended second writer or automation.

If only some files fail, group the failures by extension and pairing state. A RAW sidecar procedure does not prove embedded-metadata behavior for a JPEG, and a RAW-plus-JPEG pair can make a selection appear to refer to a different visible member. Keep the failures isolated until the distinction is understood.

Do not rename, delete sidecars or bulk overwrite the production catalog as an initial diagnosis. Preserve the smallest failing packet and its current settings. A precise reproducible case is easier to resolve than an entire assignment whose original evidence has been “cleaned up.”

A worked membership check after the packet passes

Once the six-file rehearsal works, use a bounded copied group with 36 intended candidates. Retain the 36 source identities before the handoff. The receiving Classic view also shows 36 candidates, but the comparison reveals that two intended files are missing and two other files have appeared. Equal totals have concealed four rows requiring investigation.

Separate the result into shared, missing and unexpected identities. In this example, 34 identities are shared, two are missing from the receiver and two are unexpected there. The union contains 38 identities, while each individual selection still contains 36. These are illustrative set calculations, not a claim about a normal transfer failure rate.

Inspect each exception against its original path and expected metadata. One missing photograph may have a sidecar left in another folder. Another may already be cataloged with an older rating. An unexpected photograph may be the JPEG member of a pair or a repeated basename from a different camera folder. Keep these as hypotheses until the file and settings evidence supports them.

Correct one identified cause at a time on the copy, then rerun the membership comparison. If you also rename files, change label sets and reimport simultaneously, a successful-looking result will not explain which action solved the problem. A controlled correction creates a procedure the photographer can repeat.

After identity reconciliation, inspect the required fields on the resolved set. Membership success does not prove captions, labels or development state. Likewise, complete fields on a wrong source do not make it the right photograph. The worksheet should have distinct checks for identity, selection values and important retained state.

Save the receiving selection as an explicit job group using the supported catalog workflow, then reopen the catalog. Confirm that the resolved membership remains available after the temporary import view has gone away. This tests whether the photographer has a durable working set rather than a transient screen that happened to look right.

Finally document who now owns later changes. If Classic is the next writer, make that explicit before someone resumes rating the same files in Photo Mechanic. If a round trip remains required, keep it a separate tested procedure. A clear end to the intake handoff prevents yesterday’s authoritative selection from becoming tomorrow’s competing record.

The useful outcome is not a promise that conflicts can never occur. It is a process that detects them early, preserves evidence and gives the photographer a specific next action instead of encouraging a bulk overwrite of the entire job.

What a connected workspace could remove

A shared internal catalog could avoid reimporting the same selection into a second editing environment. To earn that benefit, it must keep original identity, current decisions and historical outputs understandable. Internal integration does not remove the need for recoverable sources or exportable records.

Lightbox Pro’s reviewed build 0.1.2 (71) implements local selections and catalog organization alongside developing RAW and native export work. It does not establish Photo Mechanic/Lightroom recipe translation or general XMP writeback interoperability. Its current RAW photographic acceptance also remains open.

The longer-term aim is less repeated matching, not a claim that this existing handoff is broken by definition. A well-configured and verified Photo Mechanic-to-Classic process can be entirely reasonable. Use this packet to preserve what already works before changing it.

Official references

1. Camera Bits: IPTC/XMP write preferences
2. Camera Bits: compatibility snapshots
3. Adobe: metadata reading and conflicts
4. Adobe: ratings and color-label sets
5. Adobe: importing from existing folders
6. Camera Bits: preview behavior after editing

Product documentation checked 22 October 2026. Examples and worksheets are original teaching scenarios. Confirm the current version, camera support and terms before adopting a workflow.