PDF Page Numbers Don't Match the Printed Page? Page Labels

PDF page numbers don't match the printed page when the file has pages the printed numbering skips or restarts on: a cover, a table of contents, an excerpt that begins at page 212. The reader counts sheets from one. Page labels let you renumber that counter, and your court's rules decide which number you cite.

Why does the PDF page counter disagree with the printed page number?

The PDF page counter disagrees with the printed number because the two count different things. The counter gives each sheet in the file its position, starting at one. The printed number is whatever the author, court reporter or stamping tool put on the paper, and it can start anywhere.

Four ordinary situations open the gap.

Search the Bates number in the corner and the deposition page line in the header, then read both on the page image. Write the file page beside them. An exhibit sticker and a court reporter's number can share one sheet. A magistrate's order may show yet another printed number, so record which line you matched.

A fax cover sheet, a slip sheet between exhibits and a blank back side kept by a two-sided scan each add one more to the difference. The gap is also not constant through a file: it can grow at every separator page, so measure it at the page you care about and not only at the start.

What are page labels in a PDF?

Page labels are an optional name for each page, stored in the PDF, that a reader shows in place of the plain sheet count. With labels set, the counter can read iv, 212 or A-8 on the sheet that carries that printed number.

The PDF specification separates the two ideas. In section 12.4.2 of ISO 32000-1, the PDF specification, every page has an integer page index, and "a document may optionally define page labels (PDF 1.3) to identify each page visually on the screen or in print." It adds that "Page labels and page indices need not coincide." The specification's own illustration is a document with twelve roman-numbered pages of front matter, where the thirteenth sheet carries the label 1.

Three points follow from that section and matter when you work with a file.

• Labels sit in ranges. A range is a run of consecutive pages that share a numbering style, an optional prefix and a starting value.
• The styles are fixed. The specification lists decimal numerals, upper and lower case roman numerals, and upper and lower case letters.
• Labels are stored once for the whole document, in an entry of the document catalogue, keyed to the position of the first page in each range. They are not written on the pages.

Labels change nothing you can see on the page itself, and they leave the recognized text of a searchable PDF alone.

Setting page labels so going to a page lands on the printed number

You set page labels in a PDF editor by choosing a run of pages, a numbering style and a starting number, one run at a time. Adobe's instructions for renumbering pages in Acrobat put the command at All tools > Organize pages > Page labels.

In the dialog Adobe describes, you specify a page range and pick one of two options: "Begin new section", where you choose a style, enter the starting number and optionally add a prefix, or "Extend numbering used in preceding section to selected pages". Adobe notes that the tool "only changes the labels shown in thumbnails and page navigation; it doesn't add visible page numbers." If you use another editor, read its own help for the equivalent command before you rely on it.

A worked example

Marisol Vance holds a 96-sheet scan of a deposition excerpt: one cover sheet, a two-sheet index, then transcript pages printed 212 to 304.

1. Write down the runs before opening the dialog: sheets 1 to 3 are front matter, sheets 4 to 96 are printed 212 to 304.
2. Label sheets 1 to 3 as a new section in lower case roman numerals starting at 1, so they read i, ii, iii.
3. Label sheets 4 to 96 as a new section in decimal numerals starting at 212.
4. Check the last sheet. Its label should read 304, the number printed on it. If it reads 303 or 305, a sheet is missing or doubled somewhere in the run, and the labels have just found it for you.
5. Type 250 into the page box and confirm the page that opens is printed 250.

Step 4 is the reason to label by runs and not by guesswork: a run that ends on the wrong number is a page-count problem, and you want to know about it before anyone cites the file.

Which number do you cite from a scanned PDF: file page, printed page or Bates?

Cite the number your court's rules or the governing order tell you to cite, and say which kind it is. A scanned PDF can offer three: the sheet's position in the file, the number printed on the page, and a Bates number stamped on a production.

The rules differ by court and by type of filing, so read yours. As one example, for briefs in the federal courts of appeals, Federal Rule of Appellate Procedure 28(e) says "References to the parts of the record contained in the appendix filed with the appellant's brief must be to the pages of the appendix." The same subsection gives "Transcript p. 231" as a form of reference to an original document. Trial courts, state courts and individual judges set their own conventions.

Whatever the rule, three habits keep a citation checkable.

• Name the kind of number in your notes: "PDF sheet 14", "printed p. 221", "VANCE-000412".
• Never convert one kind to another by arithmetic alone. Open the page and read the number off it.
• When recognized text supplies the page number, read it against the picture, because a misread digit sends the reader to the wrong page.

The post on quoting text from a scanned PDF covers page and line references for a quotation, and the post on OCR and Bates-numbered documents covers keeping each page tied to its stamp.

What happens to page labels when PDFs are split or combined

Treat page labels as lost or wrong after any split or merge until you have checked them. Labels are stored for the document as a whole and keyed to page positions, so a new file made from some of the pages, or from several files, has no reason to carry the old numbering correctly.

What a particular editor does with labels when it extracts or joins pages is a matter for that editor's documentation. Adobe's page on extracting pages in Acrobat, for instance, says bookmarks and article threads tied to the pages are not extracted, and it does not mention labels either way. So test it on your own file:

1. Before restructuring, note the label and the printed number of the first and last page of every part.
2. After the split or merge, open each of those pages and compare the label with the print.
3. Reset any range that is off, one run at a time.
4. Save, reopen and check once more, since a label you see in an open session is no proof it was written to the file.

Combining is where labels pay off most, because a bundle of five exhibits has five places where printed numbering restarts. A prefix per exhibit (A-1, B-1) keeps them apart. The guide to combining searchable PDFs covers what else a merge does to the text and the bookmarks.

A searchable copy exported from Schema Law is described on the site as your original file with a hidden text layer. Check its labels as you would any new file before you cite from it; the export formats section lists what each download contains.