Word to PDF looks different? Why, and how to fix it

Why a PDF can look a little different from Word

A Word file is not a picture of pages. It holds the text, the styles and settings such as margins and spacing, and Word works out where each line and page ends when it opens the file. That is why the same document can end on a different page on another computer. Converting it means doing that same work again, and that is real work. Word to PDF does it in your browser: it reads the document, lays out every line and page, and writes the PDF itself, with the fonts inside it.

Most of the time the result is very close. We compared it with LibreOffice, a free program that also lays out Word documents (we don't have Microsoft Word on our test computers, so we couldn't compare with Word itself). For everyday documents, the pages matched: a resume, an offer letter, a leave letter, a school admission form and a Hindi office order scored 98.7 to 100 out of 100, with the same number of pages. For 22 of our 23 test documents the page count was the same.

When the PDF does differ, the cause is nearly always one of the things below. Find yours in the table.

The letters look plainer, or the lines break differently

Word's fonts, such as Calibri, Times New Roman and Arial, belong to Microsoft and can't be shared on a website. So Word to PDF draws each one with an open font made to have exactly the same letter widths. Every word takes the same room as with Word's own font, so lines and pages can break in the same places, while the shapes of the letters differ slightly. This is the list:

Other fonts get the closest of these. Verdana, Tahoma, Helvetica and similar plain fonts are drawn with Arial's widths. Georgia, Garamond, Book Antiqua and other fonts with small feet on their letters get Times New Roman's. Consolas gets Courier New's. A font it doesn't know gets Calibri's widths, unless the document says it is a serif or fixed-width font. These fonts have different widths from the real ones, so in a document that uses them, lines can break in other places than in Word, and a fancy font looks plainer.

The fix: if the exact look and line breaks matter, change the document's font to one in the table before you convert: Calibri, Cambria, Times New Roman, Arial or Courier New. The PDF then matches Word's layout most closely. Use the same font for the whole document if you can.

Aptos. Aptos is the font newer versions of Word start with. It is drawn with Calibri's widths, but Aptos is wider, so a document in Aptos may fit more words on each line in the PDF and end up shorter than in Word. If your page count matters, switch to Calibri before converting.

The PDF has more or fewer pages than Word

The page count follows the lines, and the lines follow the letter widths. With Word's usual fonts, the page counts matched in our tests: 22 of 23 documents had the same number of pages as in LibreOffice, and our 129-page book had 129 pages in both, with 127 of them starting with the same line.

Where it differed, here is what we saw:

• A newsletter in two columns made by a program came out at 3 pages here and 2 in LibreOffice. The file leaves out how its line spacing is meant to be measured, and the two read it differently.
• A resume made with a program gave two different widths for a table's columns, and the two converters used different ones, so a college name fit on one line in one and took two lines in the other.
• Lines with English words in brackets in some Indian-language notices broke one word earlier or later, so the lines after them sat a few millimetres higher or lower. The words were all there and in order.

If a page matters, such as a one-page resume that must stay one page, check the page count of the PDF. To fix a PDF that is a line too long, make the font or the spacing slightly smaller in Word, or narrow the margins a little, and convert again. A table of contents, a date or another field shows what Word saved in the file. Only page numbers are worked out again, so if pages break differently, a table of contents can be off by a page.

Grey boxes, charts, equations and old pictures

When something in a document can't be drawn, the PDF is still made, and the row says what is different. The notes you may see, and what to do:

• A chart or SmartArt diagram shown as a grey box. In Word, copy the chart, paste it as a picture (Paste Special, then Picture), save, and convert again. In our test, a report with a chart had one grey box and scored 92.6 out of 100 against LibreOffice for that reason.
• An equation written as plain text. The letters and numbers are there, without fraction bars or root signs. Pasting the equation as a picture in Word fixes it too.
• Pictures shown as grey boxes. These are EMF and WMF drawings, an older Windows format. Save them as PNG pictures in the document.
• Text in a script it can't draw yet, such as Urdu or Sinhala, which shows as empty boxes.

JPG and PNG pictures in the text, pictures placed on the page with text around them, see-through signatures, text boxes and simple shapes all convert. So do tables with borders, shading, merged cells and repeated header rows, and tick boxes in forms.

Hindi or another Indian language comes out as English letters

Hindi, Bengali, Gurmukhi, Gujarati, Odia, Tamil, Telugu, Kannada and Malayalam work, when the text is Unicode. Phones, Google's Indic keyboards and the Indian keyboards in Windows all type Unicode, so most documents are. But older documents typed in fonts like Kruti Dev, DevLys, Chanakya or Bamini are not Unicode: they hold English letters that only look Indian in that one font. In a PDF made anywhere, those come out as English letters, and the row tells you when a document uses such a font.

The fix: change the text to Unicode first. A "Kruti Dev to Unicode" converter does it. Or retype it in Mangal or Nirmala UI. The whole story is in Word to PDF in Hindi and Indian languages. Indian text in italics is drawn upright, because the Noto fonts for these scripts have no italic style, and Indian text in bold is drawn bold.

The tool won't convert the file at all

• "An older Word file (.doc)". In Word, choose File, Save As, Word Document (.docx), then convert the new file.
• "Locked with a password". Open it in Word, remove the password (File, Info, Protect Document), save, and try again.
• "Couldn't be read as a Word document". The file may be damaged, or another kind of file renamed to .docx. Try downloading it again, and check the ending of its name.
• "The fonts couldn't be downloaded". Check your internet connection and add the document again.
• "The converter stopped". The document may be too big for the device's memory. Close other tabs and apps, or use a computer. In our computer test, the browser used at most 611 MB while it converted ten documents, the 129-page book among them. A phone has less, so a long document full of big photos is where a phone is most likely to give up.

The post Convert Word to PDF without Microsoft Office has the basic steps on a phone.

A five-minute check before you send the PDF

1. Count the pages against what you expect.
2. Read the first line of each page, to see that the page breaks are where you want them.
3. Look at tables, and check the columns and the header rows.
4. Check pictures, charts and signatures are there, not grey boxes.
5. Check the names, numbers and dates on a resume or a form.
6. Click a link, if the document has links. They keep working, and headings become the PDF's bookmarks.

Keep the Word file next to the PDF. If something is off, you can fix the Word file and convert again in seconds: a one-page letter takes 0.1 to 0.2 seconds on our test computer, and a 129-page book took 1.8 to 2.2 seconds. A phone may take longer.

Your document stays on your device

Word to PDF has no upload step. Your own browser reads the document and writes the PDF, and what it downloads from this site are the converter and the free fonts, never anything about your file. The page contacts only this site and Google's ad service, which shows the ads that keep the tools free, and our code never gives your files to the ad code. The PDF doesn't copy the author's name from the Word file. The privacy policy has the details.