PDF (Portable Document Format) is a page description format: it records exactly where every character, line and image sits on a page, so a document looks the same on any screen and any printer. Invoices, tax forms, contracts, scanned letters, boarding passes, datasheets and CVs travel as PDF. If someone sent you a file, chances are it was exported from Word, LibreOffice or InDesign, or produced by a scanner.
Origin and standards
Adobe introduced PDF together with Acrobat 1.0 in 1993. The format grew through versions 1.1 to 1.7, adding encryption, forms, transparency (1.4), object streams for smaller files (1.5) and embedded 3D and multimedia. In 2008 Adobe handed version 1.7 to ISO as ISO 32000-1. Its successor, PDF 2.0, is ISO 32000-2 (2017, revised 2020); since April 2023 it can be downloaded free of charge through the PDF Association. Several subsets serve specific purposes:
- PDF/A (ISO 19005) for long-term archiving: all fonts embedded, no encryption, no external dependencies.
- PDF/X (ISO 15930) for print production, with defined colour and bleed rules.
- PDF/UA (ISO 14289) for accessibility: tagged structure that screen readers can follow.
What is inside a PDF
A PDF is a collection of numbered objects (pages, fonts, images, annotations) with a cross-reference table at the end that tells the reader where each object starts. Page content is a stream of drawing commands derived from PostScript, but without loops or variables, which makes PDF fast and predictable to display. Text and vectors are compressed losslessly with Flate; images use JPEG, JPEG 2000, JBIG2 or CCITT fax compression. Fonts are usually embedded as subsets, containing only the characters actually used.
A PDF can also carry form fields, comments, bookmarks, hyperlinks, attached files, digital signatures, password protection (AES-256 in current versions) and XMP metadata. Changes are often appended to the end of the file as incremental updates, which is how a signed PDF can keep the original signed version intact.
What PDF does not store is the logical structure of the source document. Unless the file is tagged, a paragraph is just a set of positioned lines, and a table is text placed in columns. That is why editing and converting PDFs back to Word is hard.
How to open and edit PDF files
- Windows 11: Microsoft Edge opens PDFs out of the box, including form filling and simple annotations. Adobe Acrobat Reader and SumatraPDF are free alternatives.
- macOS and iPhone: Preview and the Files app display, annotate, sign and merge PDFs.
- Linux: Okular, Evince or MuPDF; LibreOffice Draw can edit text in simple PDFs.
- Android: Google Drive or any web browser.
Real editing of text in a PDF needs Acrobat Pro or a similar editor. For larger changes it is easier to go back to the source document.
Common problems
PDF won't open or is “damaged”
Usually the download or e-mail attachment was cut off and the file lacks its end, including the cross-reference table. Download it again. Some readers can repair a broken table; MuPDF and Okular are tolerant.
PDF asks for a password
There are two kinds. An open password encrypts the content and cannot be bypassed without the password. A permissions password only restricts printing or copying in compliant readers.
Text cannot be selected or searched
The PDF is a scan: every page is a picture. It needs OCR (text recognition), for example in Acrobat Pro or with the free OCRmyPDF.
PDF file is too large
Typical causes are scans at 600 dpi in colour and photos inserted at full resolution. Re-exporting from the source with image downsampling usually helps most.
Converting PDF
Our converter offers many targets for PDF, and each behaves differently:
- PDF to DOCX, DOC, ODT, RTF: the page is rebuilt with text in positioned frames. Good for copying text, tedious for heavy editing. There is no OCR, so scans give pages with images only.
- PDF to EPUB, MOBI, AZW3, FB2: the text is reflowed for e-readers. Works well for novels, poorly for multi-column layouts and tables.
- PDF to JPG, PNG, SVG, EPS: only the first page, rendered at 150 dpi. For all pages choose TIFF (one multi-page file) or CBZ (a ZIP with one JPG per page).
- TXT and HTML for plain text extraction, PS for printing workflows, DjVu for scanned books.
PDFs protected with an open password cannot be converted. The opposite direction, for example DOCX to PDF or JPG to PDF, has its own pages. Related formats: PostScript, XPS, Adobe Illustrator.
Facts
| Developer | Adobe (since 2008 maintained by ISO) |
|---|---|
| First release | 1993 (PDF 1.0, Acrobat 1.0) |
| Current version | PDF 2.0, ISO 32000-2:2020 |
| Type | fixed-layout document, binary object structure |
| Compression | Flate (lossless), JPEG, JPEG 2000, JBIG2, CCITT |
| Typical use | exchanging and printing finished documents, archiving (PDF/A), forms |
Further reading: the Library of Congress descriptions of the PDF family and of PDF 2.0.