Most everyday PDF work is page surgery: combine a cover letter, CV and certificates into one file; pull the two signed pages out of a long contract; split a scanned stack into separate invoices; turn a sideways scan the right way up. None of these change the content of a page, only which pages exist and how they are arranged.
This guide explains how page ranges are written, the difference between rotating a page for viewing and rotating it in the file, what is and is not preserved when PDFs are merged or split, how to deal with password-protected and scanned PDFs, and how to keep file sizes and private documents under control.
How a PDF is put together, and why page tools are lossless
A PDF is a collection of objects: pages, the fonts and images they use, and a page tree that lists the pages in order. Merging, splitting, extracting and rotating mainly rewrite that page tree. The text, vector graphics and images on each page are copied as they are, not re-rendered.
That has a useful consequence: done properly, these operations are lossless. Text stays selectable and searchable, and images are not recompressed. If a tool makes your text unselectable or your scan blurrier after a simple merge, it has converted pages to images, and you should use a different tool.
Page range syntax
Almost every PDF tool, print dialog and command-line utility uses the same simple notation for choosing pages. Page numbers are the physical position in the file, starting at 1, which is not always the number printed on the page: a report with a cover and roman-numbered front matter may have printed page 1 on physical page 5.
- Most tools ignore spaces, so "1, 3-5" works the same as "1,3-5".
- Some tools accept open-ended ranges such as "5-" (page 5 to the end); many simple ones do not, so check the page count and write the end page explicitly.
- Numbers outside the document are either ignored or rejected. Verify the page count of the output.
- To reorder pages with a tool that keeps document order, extract the pieces separately and merge them in the order you want.
3 single page
2-5 pages 2, 3, 4 and 5
1,3,7 pages 1, 3 and 7
1-3,8,10-12 ranges and single pages mixedRotation: changing the view vs changing the file
Rotating a page in a PDF viewer (the rotate button in Chrome, Edge, Acrobat Reader or Preview) usually changes only your current view. Close the file, send it to someone, and the page is sideways again unless you explicitly saved the change.
Each PDF page has a /Rotate value, in multiples of 90 degrees, that tells every viewer how to display it. Changing that value fixes orientation permanently for everyone, without touching the page content, so quality is unaffected and text remains selectable. This is what a proper rotate tool does. A 180° rotation fixes upside-down scans; 90° or 270° fixes pages scanned sideways.
Mixed documents, with a few landscape tables among portrait pages, are the awkward case. If a tool rotates every page at once, extract the sideways pages, rotate them, and merge everything back in order.
What merging and splitting can lose
Pages carry over cleanly, but document-level features do not always survive, especially in lightweight and browser-based tools that work by copying pages into a new document:
- Bookmarks (the outline panel) are often dropped, because they belong to the document, not to individual pages.
- Fillable form fields may stop working or vanish, and two files with fields of the same name can conflict. Fill and flatten forms, or print them to PDF, before merging.
- Digital signatures become invalid when the signed document is changed. A merged or split copy of a signed PDF is no longer verifiably signed, so keep the original signed file as well.
- Internal links and a table of contents that jump to page numbers can point to the wrong places once pages move.
- Document metadata (title, author) usually comes from the new file, not the originals.
Encrypted, scanned and very large PDFs
PDFs can have two kinds of password. An open (user) password encrypts the file so it cannot be read without the password; no tool can merge or split it until you open it and save a decrypted copy. A permissions (owner) password lets anyone view the file but restricts printing, copying or editing, and many tools refuse to modify such files. If it is your document, remove the restriction in the software that created it or by printing to a new PDF, then do the page work.
Scanned PDFs are simply images inside pages. Rotate, split and merge work exactly the same, but the text cannot be selected or searched unless OCR has been applied. Merging scans of different sizes is fine; each page keeps its own dimensions.
File size after merging is roughly the sum of the inputs. Splitting does not always shrink files proportionally, because every output may carry its own copy of shared fonts and images. If the result is too large for an upload limit, reduce the images at source and re-create the scan or PDF, rather than relying on page tools to compress it.
A practical workflow for sensitive documents
Bank statements, ID proofs, medical reports and contracts are exactly the documents people need to merge and split, and uploading them to an unknown server is an unnecessary risk. All four Neemtools PDF page tools run in your browser using the pdf-lib library, so files stay on your device.
Start by fixing orientation with Rotate PDF. Use Extract PDF Pages to pull out just the pages you need with a range such as 2-5 or 1,3,7, or Split PDF to break a document into one file per page. Then use Merge PDF to combine the pieces in the order you want. Finally, open the result and check the page count, orientation and any links before you send it.