Sample files

Before, after and the report.

A demonstration annual report we produced from one source, not a customer file. The before file is a plain untagged export; the after file is the same document tagged and reviewed. Open them in any PDF reader, or run either through the free checker.

Before: untagged export (PDF)

12 pages, 91 KB. No tags, no language, title not displayed. This is what most authoring tools produce.

After: reviewed repair (PDF)

12 pages, 223 KB. Tagged, language and title set, bookmarks, alt text on every figure, header cells in the table.

Delivery report (PDF)

2 pages, 79 KB. The record of what was checked, when, with which tools, and what was left unresolved.

Verify it yourself

The checker runs on our own functions and shows the same findings list you would get for your own document.

What a screen reader hears

Before · untagged PDF

  • Heading read as plain text; no way to jump between sections
  • Chart image announced as “graphic” with no meaning
  • Table read cell by cell without headers
  • No language, so names are mispronounced

After · reviewed repair

  • Headings H1–H3 in a consistent hierarchy, bookmarks added
  • Chart described in a sentence; decorative rules marked as artifacts
  • Table header cells with scope; reading order verified with NVDA
  • Language en-US, title displayed, PDF/UA identifier and validation run

These are illustrations. Download the actual before file, the repaired file and the delivery report.

What changed between the two files

  • Structure tree added: one H1, 11 H2 section headings, 32 paragraphs, 3 figures and 1 table, in reading order.
  • Alternative text written for each of the 3 bar charts, stating the values shown; captions tagged.
  • Table header row tagged as header cells with column scope, so every value is announced with its column.
  • Document language set to en, title set and displayed in the window title, bookmarks added for every section, running footers marked as artifacts.
  • Content and layout unchanged: the repaired file reads the same on paper and on screen.

The delivery report

Standard
WCAG 2.1 AA with PDF/UA-1 structure
Review date
2026-09-15
Reviewer
PDF Access accessibility specialist
Tools
Adobe Acrobat Pro; NVDA 2025.1 with Adobe Acrobat Reader (Windows); VoiceOver with Preview (macOS); PAC 2024; Acrobat accessibility checker

Checks performed

Structure tree: one H1 (document title), 11 H2 section headings in order, 32 paragraphs, 3 figures and 1 table tagged; no heading level skipped. Reading order: document read start to finish with NVDA 2025.1 in Adobe Acrobat Reader and with VoiceOver in Preview; every section announced in visual order, running footers marked as artifacts. Figures: 3 bar charts given alternative text that states the values shown; captions tagged as Caption elements. Table: header row tagged as TH with column scope; each value announced with its header; no merged cells. Document properties: language en, title set and displayed in the window title, bookmarks for every section, fonts embedded. Automated validation: PAC 2024 and the Acrobat accessibility checker report no errors after repair.

Unresolved limitations

The three bar charts are described in alternative text that includes every plotted value. A separate data-table equivalent was not added because the same values appear in the surrounding text. Colour contrast of the printed chart labels was measured at 7.1:1 and needs no change.