Turning off imported metadata fields

[dkorman]

Mario,

First, as always, I find that iMatch continues to amaze me in its flexibility and attention to detail.

I use iMatch with many non-image files, which it handles extremely well, but I would like to limit certain imported metadata fields, for example keywords and description fields.  I have tested using the Tags option (under Preferences->Metadata2) unsuccessfully on PDF files.  If iMatch originally imports the metadata, even after I change the PDF and XMP tag settings (even unchecking them all) and then try various combinations of Database Diagnosis (which indicates that the Tags settings have changed), Reloading Metadata and Forced Updates, the Metadata Panel and Keyword Category still show the original loaded metadata.  Am I doing something wrong?  Also, what Tag(s) handle EPUB files (whose metadata I generally find to be a complete mess)?

Thank you.

[Mario]

These settings only affect files you add to the database afterwards.
XMP data will always be imported / created by IMatch and ExifTool.
A few hundred thousand metadata records don't affect IMatch. If you do this for performance reasons to to make your database smaller...unless you have tens of thousands of PDF files, you would not even notice.

[dkorman]

Mario,

I actually do have, in some databases, non-image files in the many tens of thousands, which contain metadata which is so messy and useless to me that I would like to exclude it.  For me, at least, I believe that I might be adversely affected by the superfluous data: affecting file size, performance (e.g. the continuous requirement to refresh Keywords), overhead when Rescanning, as well as search performance.  Further, at least in the case of Keywords, I know that certain file types may have useful Keywords, but Keywords from PDFs, EPUBS, etc. overwhelm the good with the bad. 

Your writing that the Tab setting only affects subsequent imports seems to be at odds with the documentation, unless I misunderstand it.  Specifically, when I make changes to the Tag settings (for example, unchecking all the PDF and XMP Tags), I get the following message on exit:

"If you excluded groups or tags, the data associated with these groups and tags remains in the database until you run the next Database Diagnosis.  If you have included new groups or tags, you need to rescan files (or reload metadata) to import the data for these tags into the database."

Then, when I run Database Diagnostics, the message appears:

"If the database contains data for tags which are no longer imported, this extra data will be removed during the diagnosis.  This reduces the amount of data stored in the database and improves performance."  with an option to "Remove superfluous data now.".

If iMatch cannot currently exclude certain fields by file type (which is what I assumed Tags editing was designed for), then I would request adding this as a feature to your next version.

Thank you as always.

[Mario]

I remembered that wrong. Did not use this for years.
IMatch will remove unmapped tags, and the diagnosis even reports how many tags have been removed.

[dkorman]

Thank you for your quick response, but I'm still in the dark. 

1) Can I turn off the import of certain fields (e.g. Keywords and Subject/Description) for PDFs and EPUBs (or other file types)?  If so, how, because when I tried it for PDFs, it didn't seem to make any difference.  There does not appear to be a Tag for EPUBS (which are basically zipped HTMLs).

2) Can I remove fields from an existing database for PDFs and EPUBS (or other file types) from an existing database?  If so, again, how, because I could not get it to work as documented.

Thanks.

[Mario]

Keywords and description are part of XMP. You cannot remove XMP from your database, it's the core of IMatch's metadata management.
For other tags, you can remove them via the Tag Manager.
Just tried with some Nikon maker notes, works as documented.

[dkorman]

Mario,

So, I now (somewhat) understand iMatch's capabilities in this area.  Again, maybe this could be a feature in the next version.

Thank you for attending to my questions.

[Mario]

Again, maybe this could be a feature in the next version.

What could be a feature? Removing XMP data?
Nope, sorry. That's the core of IMatch's metadata management and the best cross file format metadata standard available today.