ocrmypdf produces wrong page size

[femifrak]

In contrast to
ocrmypdf in.pdf out.pdf
ocrmypdf --force-ocr in.pdf out.pdf
produces an output page format (115 × 200 mm) different from the input (A5, 148 × 210 mm).

I've been using pikepdf 8.14.0, ocrmypdf 16.4.1 / Tesseract OCR-hOCR 5.4.1.

in.pdf
out.pdf

[femifrak]

The behaviour that-fmay change the output page size appeared first with version 16.1. (16.0.4 does not show this bug.)

This is true for both renderers (sandwich and hoc).

in.pdf is a simple file without text but the same effect happens in pdfs with text: pages will be cut off in the middle of the text.

[Jmuccigr]

Is it perhaps this bug? #1181

[femifrak]

In my case it was sufficient to use --redo-ocr instead of --force-ocr. --redo-ocr does not have that issue.