Option to retain original page dimensions?

[someone-i-know]

I tried OCR for a pdf in which each page is roughly A4 dimensions.. The OCRed file has large portions of empty space around the text, increasing the page width and height. Is there a way to retain the original page height and width while performing OCR?

I cannot show the pdf but this is what it looks like: (left is OCRed, right is original)

[jbarlow83]

Generally speaking ocrmypdf preserves original page.

This issue concerns a specific PDF file and I will need that file to test it.

If the file contains private information, here are some options:

• You can encrypt the file with my public key as described here, to ensure that only @jbarlow83 can open it.
• You can attempt to produce a non-private file that likely has a similar issue. For example, if the PDF that caused the problem was created with a specific scanner software, perhaps that software is the problem. Scan some other document or a redacted document with the same equipment and software. (If you are not sure what to scan, print a page from Wikipedia.)
• You could try using Acrobat's Redact tools, or similar tools in other PDF editors, to remove personal information. Note that this will change the PDF and might fix the issue.
• You could run qpdf --json yourfile.pdf > yourfile.pdf.json and upload the resulting .json file here. That contains a structural description of the contents of the PDF. In most cases, this will not help; but occasionally it does.

If you cannot share the file, you could try the following workarounds to produce a visually similar PDF:

• Refrying the PDF using Ghostscript gs -q -sDEVICE=pdfwrite -o out.pdf in.pdf.
• Printing the file to a "Print to PDF" virtual printer, which will reconstruct the PDF and perhaps create a new PDF without the issue.
• Using ocrmypdf --force-ocr which is more aggressive about removing document contents.

If you still cannot share the file after the options above, please close the issue, because there is no way I will be able to reproduce it if you cannot provide a way for me to test it.

[someone-i-know]

Thanks for responding. I have extracted one page from the file which exhibits this same behaviour.
I hope this helps to identify the problem.

to my mother.pdf

Command used:
ocrmypdf --force-ocr [input path] [output path]

[jbarlow83]

This is a bug - will address soon

[jbarlow83]

#1181