What Extract Pages Preserves and Changes
Extracting pages produces a new document. In Extract Pages V1 some things come across with the selected pages and some change; this guide lists which.
Knowing which is which takes a minute and saves sending someone a file whose links go nowhere.
In short
Extract Pages copies selected pages into a new PDF without converting whole pages to images. It does not perform OCR, so scanned pages do not become selectable text. Web links are retained, but links to pages inside the original PDF are removed. Form fields become fixed page content, and bookmarks are not carried over. A digital signature on the original does not validate the new file. The original PDF is left unchanged.
What comes across unchanged
The page itself. Extract Pages copies the selected pages and the resources needed to display them into a new PDF. Pages are not converted to images as a whole. Existing selectable text remains selectable. Extract Pages does not add OCR to image-only scans.
Page geometry. Page size, rotation and the page’s boxes come with it. A page whose visible area differs from its paper size — a cropped scan, a trimmed print layout — keeps both, so the new file shows the same area the original did.
Document information. Title, author, subject, keywords and creation date carry over as they were. Extract Pages does not add “(extracted)” to the title; which pages you took is something for the file name.
Links to web addresses. A link pointing at a URL still works in the new file.
What changes
Internal links. Extract Pages V1 removes links to pages inside the source PDF, including links whose destination page is selected. It does not rebuild those destinations in the new document. Links to web addresses are retained.
Form fields. Extract Pages V1 turns form fields into fixed page content. Their visible values remain, but the fields can no longer be filled in or edited as form controls. If this conversion fails, the app does not produce a file.
Bookmarks. Extract Pages V1 does not copy or rebuild the document’s bookmark navigation.
Structure that belongs to the document. A few things are properties of the whole document rather than of any page, and a new document does not inherit them: the tags that let a screen reader read the file in the right order (its accessibility structure), XMP metadata, files attached to the document, named destinations, layer definitions and custom page labels. Page content itself is unaffected. If the file has to stay accessible, send the original or tag the extracted file again in a tool that can.
These are limits of the current version, not a claim that every PDF extraction tool must behave this way.
Digital signatures. The original document’s digital signature does not validate the new PDF. Keep the original signed file when signature verification matters.
Password-protected files. Encrypted PDFs are not opened at all — neither files that ask for a password nor files that open freely but restrict printing or copying. There is no field to enter a password, and the page says so rather than opening the file halfway.
What Extract Pages checks
“Byte-for-byte identical” is not something extraction promises. The output is written fresh: object numbers, the cross-reference table and compression belong to the new file.
What Extract Pages does check is narrower: before the file is saved, it is reopened and each page’s drawn content, size and rotation are compared with the page it came from. If that check fails, no file is produced.
Nor is extraction a way to redact. The pages you keep carry everything drawn on them, and the document information comes across as it was, so leaving pages out does not clean a file of hidden or sensitive data.
Checking before you send
Read the list of pages before saving rather than the count, open the result once, and if the document had forms or internal cross-references, look at one of them.
Doing it in Extract Pages
- Open the PDF. If it is signed, has a form or has internal links, that is said on screen before you save.
- Pick your pages and read the Pages field.
- Save, and open the result once.