Copy link to clipboard
Copied
No filter must be specified explicitely for metadata objects. Metadata must be embedded in clear text so non-PDF-aware applications can extract the XMP metadata.(c) ISO 19005-1:2005
How to remove the filter in the metadata of a PDF? AcrobatPro v.10.
Copy link to clipboard
Copied
This is specifically for PDF/A compliant files, and does not apply to normal PDF documents. When you convert your document to a PDF/A file, then the metadata should be saved correctly. Other than that, I am not aware of another way to force the metadata dictionary to be saved uncompressed from within Acrobat. [ Update: It can actually be done using the PDF Optimizer, see further down for more information ]
Are you running into problems with PDF/A files and compressed metadata dictionaries?
Copy link to clipboard
Copied
Yes, I need create a correct PDF/A file, but when I save the filters remain in the metadata.
Copy link to clipboard
Copied
How exactly are you creating PDF/A files? Which PDF/A "flavor" are you using? I just did a quick test, and when I use the PDF/A profiles in Preflight, a compressed metadata stream is correctly uncompressed and saved in the resulting PDF/A file.
Copy link to clipboard
Copied
I use "Convert to PDF/A-1b (sRGB)" in Preflight, click on "Analyze and fix". After conversion show a message: "No problems found", PDF/A file created, but the filters contains in the metadata.
When I try to use this file, an error occurs:
No filter must be specified explicitely for metadata objects.
Copy link to clipboard
Copied
In "Internal PDF structure" there is a string: "Filter: FlateDecode". If I understand correctly, the problem in compression.
But remove the compression settings in "Optimize PDF" does not solve the problem.
Copy link to clipboard
Copied
This is the content stream of a page.
Copy link to clipboard
Copied
As Bernd explained, you are looking at a page content stream, which can be compressed in a PDF/A file. You need to look in the document's "root" object for the "Metadata" stream:
As you can see, I my Metadata stream is compressed using the "FlateDecode" filter. When I use the PDF Optimizer, it does actually remove the compression from the Metadata stream:
After I run that, this is what the Metadata stream looks like:
Copy link to clipboard
Copied
Problem with filters disappears when I remove compression, but when converted to PDF/A error occurs again because return a string: "Filter: FlateDecode".
Perhaps this makes "Recompress LZW as ZIP"? It can be removed from the process of creating PDF/A?
Copy link to clipboard
Copied
This message is about other compressed content. You are interested in the line that says "Make document XMP Metadata compliant with PDF/A-1 (2 objects)".
When you validate your PDF/A document does it pass validation? Again, compressed content is fine in a PDF/A file as long as it's one of the supported compression methods. Page content streams are usually always compressed to save space.
When you check for compression, make sure you only look at the metadata, not anywhere else in the document.
Copy link to clipboard
Copied
We won it! To create a PDF/A-1 (filters free) appropriate ISO 19005-1:2005 need:
1. As Karl said, to use the PDF Optimizer, it does remove the compression from the metadata;
2. Remove the tick "Apply fixups" in Preflight (it contains "Recompress LZW as ZIP").
Thank you very much, Karl!
P.S.: Sorry for my English.
Copy link to clipboard
Copied
This may create invalid PDF/A-1 files: LZW compression is not allowed in these files. That's why that option is checked so that anything that is LZW compressed will be changed to Flate (or ZIP) compression. It should not change the compression state of the metadata stream.