Overview
If you do client-side OCR PDF conversion related to JPG/JPEG files in M-Files Desktop, in some cases it rotates pages and original and converted images are not identical.
Solution
This is known issue CAT-220. You can track it through our release notes.
Workaround
There are three workarounds:
- Workflow and server-side PDF conversion: https://userguide.m-files.com/user-guide/latest/eng/State_actions.html
or
- Client-side PDF conversion: https://userguide.m-files.com/user-guide/latest/eng/making_pdfs_of_existing_objects.html
or
- Do not invoke the conversion for image-only images.
The prompt to 'make the image searchable' (which is the OCR action) is helpful only for scanned texts (documents), it will not do anything for image-only files and may lead to this issue.
