Stirling PDF crops the wrong area of a scanned PDF

You select an area in a scanned PDF, but the cropped result removes the required content or keeps a different region. An archived original report names Stirling PDF 3.2.2 and this mismatch. Rotation and coordinate interpretation are useful hypotheses to test; the report does not establish a universal cause. Q1

The saved output is the decisive observation

The preview shows your selection, while the saved document shows the operation's actual effect. Stirling describes cropping as selection followed by applying the operation. Q2 A correctly positioned rectangle does not establish that the output retains the correct area.

Build a reproducible comparison

1. Separate original and working copy. Preserve the unmodified scan. Record scanner or export software, Stirling version, page count and affected page size. Begin with one non-sensitive page rather than a whole confidential document.
2. Use unmistakable orientation. Choose a page with distinctive content in different corners or create a dedicated test page. Select an asymmetric region. A nearly empty margin makes an incorrect shift difficult to identify reliably.
3. Record selection and output. Capture the selection view and open the saved PDF in an independent viewer. Compare retained text, visible corners and orientation. This distinguishes a preview problem from an incorrectly generated file.
4. Compare page characteristics. Determine whether the problem affects only rotated pages, a particular scanner output or mixed page sizes. If another export variant works, change one property at a time. A suspected CropBox or rotation issue remains a hypothesis until demonstrated.
5. Check the required page types. Do not generalise one successful test to a document containing several page formats. Inspect the result for each relevant group before sending the document onward.

Evaluate the cost of conversion

Re-exporting or normalising a copy can provide a useful comparison. Check the effect on searchable text, OCR, links, quality and existing signatures. Do not automatically replace the master with a converted file merely because its crop looks right. If the objective is permanent removal of sensitive content, use an appropriate redaction process with its own verification; cropping alone does not establish that result. Q3

For handoff, prepare a small anonymised sample, a selection screenshot, an output screenshot and the software versions. You need not disclose full confidential pages if a synthetic test page reproduces the same behaviour. Preserve the steps and settings so that an updated version can be assessed with an equivalent test.

Frequently asked questions

Will rotating the page before cropping fix it?

That can be a comparison, but it is not a verified repair until the exported result is checked.

Is cropped-out information permanently removed?

A visible crop is not proof of permanent removal. Use a dedicated redaction workflow for confidential information.

Describe your case