Page setup, sharpness and other conversion details

Where the page size and margins come from

A Word document can be split into sections, each with its own page setup. Reports often use this to put one wide table on a landscape page in the middle of a portrait document. The converter reads the page size and margins from the document's last section only, and uses them for every page. So that landscape table page comes out portrait, at the same width as the rest. Use Orientation under More options if the whole PDF should be landscape, or copy that section into its own document and convert it on its own.

A few more details about the page setup:

• Choosing A4 or Letter changes the paper but keeps the margins from the document.
• The Margins options set all four sides to the same value. Normal is 72 points (1 inch) and Narrow is 36 points.
• A margin wider than a third of the page's shorter side is reduced to that, so a document with odd settings can't squeeze the text into a thin strip.

How sharp the pages are, and why it depends on your device

Each page is drawn in the browser and stored as a picture, at 1.5 times your screen's pixel density, capped at twice. On an ordinary laptop screen that works out to about 144 pixels per inch. On a phone or a high-density laptop screen it's about 192. The same .docx converted on a phone therefore comes out sharper, and larger.

Pages are saved as PNG, which keeps text edges clean and white paper pure white. When a page is mostly photos and the PNG would be big, the tool switches it to JPEG at 92% quality if it's less than half the size. If the finished file is too big to send, Compress PDF can shrink those page images further.

Why the text sometimes comes out tiny

A table too wide for the space between the margins isn't cut off. The converter lays it out at its natural width, up to 2,400 pixels, and then scales the whole document down to fit the page. That keeps everything visible, but the text gets smaller everywhere, not only in the table. If a document comes out with tiny text, look for one oversized table. Switching Orientation to Landscape or Margins to Narrow gives it more room, and splitting the table in Word fixes it properly.

Links, comments and the text layer

Hyperlinks look right, blue and underlined, but they aren't clickable in the PDF, because the page itself is a picture. If readers need to click links, export the PDF from Word or Google Docs instead.

Word comments aren't included. If the document has tracked changes, accept or reject them in Word before converting, so you know exactly which text you're getting.

The invisible text layer is placed line by line over the picture. Search works and so does copying, but copied text usually comes out with a line break at the end of each line rather than as flowing paragraphs.

The PDF's document title is left empty. If the file is going into a document system that shows titles, add one with Edit metadata.