What changes when an ODT, ODS or ODP file becomes a PDF?

OpenDocument is a *live* format. A PDF is a *finished* one. Everything that changes in between follows from that.

What is kept

The text, as real selectable text — not a picture of it, so it can still be searched and copied.

Images, at the resolution they were placed at.

The layout, exactly as it appeared on the machine that did the export. This is the whole point: the recipient sees what you saw.

What is flattened

Styles become appearance. "Heading 1" in an ODT is a style: change its definition and every heading in the document changes. In the PDF it is 16pt bold text at a coordinate. It looks the same and it is no longer connected to anything.

Spreadsheet formulas become their results. An ODS cell containing =SUM(B2:B40) exports as the number it currently shows. The formula is gone; the number stays. That is usually what you want for distribution and always a surprise if you expected to keep working in it.

Presenter notes and animations, in ODP. A PDF has no timeline. Slides become pages, transitions vanish, and notes are either omitted or exported as a separate layout depending on the option chosen.

Tracked changes and comments. Depending on the export setting these are either burned into the page as visible marks or dropped. Check before sending anything sensitive — a document exported with tracked changes visible has published your edit history.

What can go wrong: fonts

This is the failure people actually meet. If the document uses a font that is not embedded, the PDF has two choices: embed it, or reference it and hope the reader's machine has it. When it does not, the reader's viewer substitutes something else, and substituted fonts have different widths — so line breaks move and a carefully fitted table stops fitting.

Exporting with fonts embedded avoids it entirely, at the cost of a somewhat larger file. It is almost always worth it. See why fonts change for what that looks like when it goes wrong.

What to check after exporting

1. Page count. A spreadsheet that was one screen can be nine pages — see content that does not fit the page width.
2. The last page. Trailing empty rows in a spreadsheet often produce blank sheets.
3. Select some text. If nothing highlights, something rasterised the document.
4. The fonts, if the layout matters. Open it on a different machine.

Converting the other way — from PDF back to an editable format — is a much harder problem, because the structure was thrown away here. That one is covered in why converting back is never perfect.