Preserve Selected Music Tags in Picard
At a glance
MusicBrainz Picard allows users to preserve selected tag values during metadata updates by specifying exact tag names in its advanced preservation settings. This approach lets Picard update release-related tags while maintaining personal or local metadata fields. To ensure accuracy, test changes on small file copies before batch saving, and verify preserved and updated tags thoroughly.
In this guide
You can preserve selected original tag values in MusicBrainz Picard while still updating other fields. This is a Picard tagging workflow, not an audio editing task. The practical approach is to identify the exact tag names you want to protect, enter only those names in Picard’s documented advanced preservation option, and test the result on a small copy of files before saving a larger library.
What preserving tags means
Preserving selected tags is different from refusing all metadata changes. The goal is usually to let Picard update release-related information while leaving a few personal or library-specific values untouched. Those values may have been added by your player, another tagger, or your own workflow.
This is useful when a field has local meaning in your collection. One listener may use genre for broad listening moods rather than release genres. Another may store comments that explain where a recording came from. A DJ may depend on ratings or grouping-like fields maintained outside Picard. If those values change, the files may still be identified correctly, but the collection becomes less useful to its owner.
The important detail is that Picard needs tag names, not vague descriptions. “Keep my notes” is not enough. You need to know the tag field name as Picard shows it, then use that name in the preservation setting. Because tags can vary by format and by the software that wrote them, treat this as a careful workflow rather than a one-click cleanup.
Relevant Picard tag options
Picard has documented tag options for writing metadata, preserving file timestamps, clearing existing tags, keeping embedded images when clearing, removing certain tag types from FLAC or MP3 files, fixing empty FLAC seek table blocks, preserving named tags from being cleared or overwritten, and overriding date sanitization for selected formats.
For this article, the central setting is Preserve these tags from being cleared or overwritten with MusicBrainz data. It is an advanced option. Use it only after identifying the fields you want to protect. A misspelled, incomplete, or mismatched field name may not protect the value you intended.
Also check whether Picard is set to write tags to files. If writing tags is disabled, Picard will not write metadata changes to the files, although other configured actions may still apply. If Clear existing tags is enabled, the task becomes more aggressive, because existing metadata is removed unless it is covered by documented preservation behavior or another relevant option.
Choose which values to protect
Before changing settings, make a short list of the tag values that matter to your collection. Avoid preserving fields just because they exist. The more fields you protect, the more likely you are to block useful updates or create confusing combinations of old and new information.
Good candidates are personal, local, or maintained by another program. Examples include ratings, comments, custom grouping information, or a genre scheme you curated manually. Weaker candidates are fields you expect Picard to improve, such as album title, track title, artist credit, track number, or release date. If you preserve those, you may prevent Picard from correcting the metadata you wanted it to manage.
Use these questions to narrow the list:
- Is this value personal or meaningful only inside my library? If yes, it may be worth preserving.
- Would MusicBrainz likely provide the better public release value? If yes, avoid preserving that field unless you have a specific reason.
- Can I verify the exact field name in a small sample? If not, do not batch-save yet.
If your wider goal is release matching rather than tag preservation, review related Picard choices first. Release selection preferences can influence the metadata Picard proposes, so you may want to read Set Preferred Release Types in Picard before changing many files.
Safe small copy workflow
The safest method is to investigate first, then test on copies. Do not begin with an entire music folder. Pick representative files: one with a rating, one with comments, one with artwork if artwork matters to you, and one file from each format you commonly manage.
- Make a working copy. Copy a few music files to a temporary folder. Leave the originals untouched while you learn how the settings behave.
- Open the copies in Picard. Review the existing metadata shown for the files and note the fields you want to preserve.
- Identify exact tag names. Use the names Picard displays for the relevant fields. Do not rely only on what another player calls them.
- Enter only the chosen tag names. Keep the preserve list short and intentional.
- Avoid broad clearing unless you need it. For preserving selected values while updating other fields, clearing all existing tags can add risk and extra verification work.
- Save only the copies. After saving, inspect the same files again and compare the protected values with your notes.
- Expand gradually. If the copy behaves correctly, try a slightly larger sample before using the setup on a major folder.
This conservative workflow helps you find a wrong field name, a format difference, or a setting conflict while only duplicate files are affected.
Example: personal genre and comments
Example: Maya has a folder containing live recordings and official albums. Her music player uses a comment field for notes such as “radio rip cleaned later” and “alternate cover preferred.” She also uses genre values such as “Sunday mix” and “gym rotation,” which are personal categories rather than release genres. She wants Picard to identify albums and update release-related fields, but she does not want those comments or personal genre values replaced.
Maya copies six files to a test folder. In Picard, she checks how the fields appear for those copies. She decides to preserve the comment field and the genre field because both represent her own library organization. She does not preserve album, artist, track title, or date, because those are fields she wants Picard to update.
She enters the relevant tag names in Picard’s advanced preservation setting, saves only the copied files, and then reviews them. The result she wants is simple: release metadata updated where appropriate, while her comments and personal genre values remain as they were. Because the sample worked, she repeats the process with one album folder before moving on to more files.
Clearing tags without losing what matters
Clear existing tags deserves special care. It removes existing metadata and leaves files with MusicBrainz metadata. That can be useful in some cleanup workflows, but it is often more aggressive than necessary when your main goal is to update identified release information while keeping selected personal fields.
If you do use clearing, the preserve list becomes more important because it is the documented way to protect named tags from being cleared or overwritten. Embedded images are separate: when clearing tags, Picard has a documented option to keep embedded images. If artwork matters, verify image behavior on copies rather than assuming the result.
Format-specific options should be treated as separate decisions. Picard documents options for removing ID3 tags from FLAC files and APEv2 tags from MP3 files, while still writing the recommended tag types for those formats. Another documented option can remove empty FLAC seek table blocks when saving. These are not general “preserve my personal metadata” controls, so do not enable them casually as part of a preservation job.
If you are experimenting with Picard automation as well as tag preservation, keep the tests isolated. A script can change tags before saving, which makes it harder to tell whether a preservation setting or a script caused the final result. For a cautious approach, see Test a Tagger Script on Music Copies in Picard and keep script testing separate from preservation testing.
Verify before batch saving
Verification is what makes tag preservation dependable. After saving your small copies, check both sides of the change: the fields you wanted Picard to update and the fields you wanted Picard to leave alone. If the preserved fields are correct but the album match is wrong, fix matching first. If the match is correct but a personal field changed, revisit the exact tag name in the preservation list.
Record what you checked:
- Which copied files were used for the test?
- Which exact tag names were entered in the preserve setting?
- Which personal values should have stayed unchanged?
- Which release fields should have been updated?
- Did artwork behave as expected, especially if clearing tags was used?
- Were results consistent across the file formats in the sample?
Keep a short note explaining why each protected field is protected. That helps later if you change computers, revisit Picard settings, or troubleshoot a folder that behaves differently.
Your next action: choose three to six copied files with the personal tags you care about, identify the exact tag names Picard shows, add only those names to the preserve setting, save the copies, and verify the results before touching a larger folder.
When adding another song to your library, a title-and-artist search on Mp3Juice can help you find the recording. Keep a backup of its existing tags before applying your Picard settings.
