Help centre

Text looks wrong in the PDF or on screen

Boxes instead of letters, Word files that lose images, plain text in other scripts and soft-looking pages: causes and fixes.

Letters that show as boxes, a Word file that lost its pictures and a soft-looking page all have a cause in how SignSpry builds the PDF. Find yours below.

What you see

  • Boxes or gaps where a name, address or stamp should be on the signed PDF or the certificate of completion.
  • A Word (.docx) file whose images, charts or text boxes are missing, or whose page breaks fall in new places.
  • A plain text file whose non-Latin text is wrong.
  • A very large page that looks slightly soft on screen.
  • Anchor tags that don't place fields, and search that finds no words, on a scanned PDF.

Why

SignSpry draws field values, stamps and the certificate itself, with Noto Sans fonts chosen per script. They cover:

  • Latin, Cyrillic and Greek
  • Devanagari, Bengali, Gurmukhi, Gujarati, Oriya, Tamil, Telugu, Kannada and Malayalam
  • Arabic, Hebrew and Thai
  • Sinhala, Khmer, Lao, Myanmar, Georgian, Armenian and Ethiopic
  • Chinese, Japanese and Korean, from a system font

Arabic, Indic scripts and Thai are shaped so letters join and vowel marks move correctly. Right-to-left text is put in the right order, and copying text out of the PDF gives the logical string. If a script has no bundled font, SignSpry falls back to Noto Sans and only that script shows as boxes.

Word files are converted by reading the main text of the document. Headings, paragraphs, lists, tables, bold, italic, underline, strikethrough, colour, size and alignment come across. Images, charts and text boxes are skipped. The result is laid out on A4 pages by SignSpry, so page breaks can differ from Word's. Plain text is set in Helvetica, a Latin-only font, on A4 pages. Line breaks and indentation are kept, a tab becomes four spaces, and one very long unbroken line is wrapped.

Page images shown on screen are capped at about 8 million pixels, with a longest side of 4,000 pixels. Ordinary A4 and Letter pages are never affected. The document itself is untouched.

Text is read from the PDF's own text layer. SignSpry does not run OCR, so a scan has no text to find.

Fix it

Boxes where letters should be

  1. Note which script the boxes replace and which field or page they are on.
  2. Open a ticket from Support in the account menu (or at Support) with the envelope title and a screenshot. A script without a bundled font needs a font added on our side, so this is not something you can fix in your document.

A Word file lost content or moved

  1. In Word, choose Save as and pick PDF. Check the PDF looks right.
  2. Accept or reject tracked changes first so the PDF shows the text you want signed.
  3. Upload the PDF. A PDF is used exactly as it is.

Plain text in a non-Latin script looks wrong

  1. Paste the text into Word or your editor and export it as a PDF, or save it as .docx.
  2. Upload that file instead of the .txt.

A page looks soft

  1. Choose a smaller page size in the program that made the PDF (for example A3 rather than a poster size).
  2. Export again and upload.
  1. Run your scanner's text recognition, or an OCR tool, and export a searchable PDF.
  2. Upload it. Or place the fields by hand on the prepare screen.
To check that a sealed PDF is exactly what SignSpry produced, drop it on Verify.
Was this useful?

More in Fix a problem

The signer says no email arrivedWhat the red "never reached its recipient" banner means, why an address is suppressed, and how to get the link to the signer. The signer can't open the linkExpired, used, replaced, not their turn, access code or one-time code: what each page means and how the sender fixes it. My file was refusedEvery upload message explained: wrong file type, password-protected PDF, size and document limits, and Word files that won't convert. A plan or pack payment failedWhat to do when a subscription payment fails, a checkout shows an error, or signers can't pay you through a payment field.