Skip to main content
BlogNaming formats

Add AI Descriptions Without Losing Original File IDs

Keep camera numbers, export IDs, or selected parts of the original filename while adding a description. Use $oldName and $words format patterns.

Published: 2 min read
On this page
The format editor lets you combine original-name tokens with the generated description.
The format editor lets you combine original-name tokens with the generated description.

Keep the part that another workflow depends on

A camera number or export ID may link a file to an edit, order, or external record. Replacing the entire name with a description can remove that useful reference.

RenameClick format patterns can retain the original stem or selected words and add the AI description. They operate on the name’s text; they do not discover the meaning of an arbitrary internal ID.

Keep the complete original stem

With the built-in RenameClick instructions selected, create a custom Format pattern. $oldName inserts the original filename without its extension; $1 inserts the generated description.

$oldName_$1

The underscore separates the two parts. The app preserves the original extension. This is a pattern to enter, not a promise of a particular generated description.

Use the full stem when every part is useful. If original names already contain descriptions, appending another may create an unnecessarily long name.

Keep selected parts with $words

Use $words{old,...} to select words from the original name. Word positions are counted from the start; use last for the final word; negative positions are not supported. The reference documents ranges, separators, and missing-word behavior.

$words{old,1}_$1
$words{old,1..2}_$1

These patterns retain the first word or the first two words, then add the AI description. Word splitting follows the configured separator rules; punctuation in an ID can divide it into several words.

Check the preview against the names you actually have. A field that is third in one export scheme may be second in another. A missing requested word is not reconstructed from the content.

See word extraction syntax and examples for the full reference.

Separate batches with different naming schemes

Apply one extraction rule only to files with the structure it expects. Use separate batches or presets when camera names, downloaded names, and export IDs follow different schemes.

Prefer $oldName if selecting a fragment would risk losing part of an identifier. Use Find & Replace on the suggestions for a consistent correction you can inspect before applying.

Do not replace an ID with a counter

An index token assigns a sequence in the current processing or apply operation. It does not preserve the camera’s number or an external export ID. Use original-name tokens for identity and counters only when you need an additional ordering number.

See index and success-index behavior for the full reference.

Check before applying

Compare the preserved part with the original name, especially punctuation, leading zeros, and mixed letter-number IDs. Check that separators are not repeated when a field is absent.

Select and apply the names only when the retained identifiers are correct. For a custom naming prompt, ordinary Format is disabled; choose the built-in instructions to use these deterministic patterns.

See planning a reusable naming convention for the full reference.