Use this checklist to self-verify multilingual manual layout before release. It applies to output from manual translation, technical document translation, and user manual translation alike. It covers page flow, text handling, typography, visual elements, and multilingual behavior such as RTL mirroring and CJK line breaking. Run it in two stages: a self-check by the DTP operator, then a cross-check by an independent reviewer.
- Who this is for Technical writers, manual owners, localization and DTP staff, quality reviewers, and manufacturers producing multilingual deliverables with automated or AI-assisted layout tools.
- When to use it During multilingual DTP or immediately after it, right after the PDF or HTML build, and at final layout sign-off.
- Use with The companion DTP and Layout Integrity Guide explains the rationale and application of each check.
These are the general, industry-neutral checks selected from the multilingual DTP standards Hansem Global has built across 36 years of localization programs for global manufacturers. They do not replace a customer document standard, a brand guideline, a published standard, or a regulation.
Document and sign-off record
| Product / model | Document version | ||
|---|---|---|---|
| Target languages | Output formats (PDF / HTML / print) | ||
| Layout tool / version | Final build date | ||
| Self-check by / date | Reviewer check by / date | ||
| Release decision | ☐ Approved: all 10 Grade A items closed at zero ☐ Rejected: return for correction and re-verification | ||
Severity grades
- A Critical / mandatory A miss can delay launch, force a full re-layout, trigger a complete reprint, or create legal exposure. Must be zero at sign-off.
- B Major / recommended A miss leads to customer claims, rework, and erosion of brand trust. Target zero wherever possible.
- C Minor / supporting Low impact on its own. Accumulated, it drives quality variance and operating cost. No Grade C items appear in this part.
- Work the items in grade order: A first, then B.
- Run two stages: a self-check, then a cross-check by someone who did not do the layout.
- Mark any item that does not apply as N/A and record the reason in Notes.
- The 10 Grade A items must be at zero before the deliverable ships.
The 20 checks
| Category A · Page flow and structure | |||||
|---|---|---|---|---|---|
| No | Check and acceptance criterion | Grade | Self | Rev. | Notes |
| 1 | Are running heads, running feet, language bars, and page numbers positioned consistently across languages? Check position, size, and alignment per language, plus the blank-page policy. | A | ☐ | ☐ | |
| 2 | Do chapters and sections start on the same side (recto or verso) in every language? Confirm the recto-start or new-page-start rule holds across all languages. | B | ☐ | ☐ | |
| 3 | Have the generated TOC and index been rebuilt in every language, with page numbers that match? Translation length shifts page numbers. Rebuild, then verify the TOC against the body text. | A | ☐ | ☐ | |
| 4 | Does recto and verso pairing hold per language in the bound version? Confirm that left-page and right-page positions stay identical across languages. | A | ☐ | ☐ | |
| 5 | Have master page changes been synchronized to every language deliverable? Confirm master page and running-head variable changes propagated to all language files at once. | B | ☐ | ☐ | |
| Category B · Text handling | |||||
|---|---|---|---|---|---|
| No | Check and acceptance criterion | Grade | Self | Rev. | Notes |
| 6 | Is there any text box overflow caused by language expansion? Check every expanding language (German, Russian, Finnish) in full, not by sampling. Zero clipped characters, zero overflow. | A | ☐ | ☐ | |
| 7 | Are hyphenation and line-breaking rules applied per language? Confirm hyphenation is turned off for languages that do not hyphenate (Chinese, Japanese, Korean, Thai, Arabic). | A | ☐ | ☐ | |
| 8 | Are abbreviations, units, and figures protected from splitting at line end? Apply non-breaking spaces so that “64 GB” and “Fig. 3” cannot split across a line break. | B | ☐ | ☐ | |
| 9 | Is paragraph alignment (left, right, justified, centered) consistent per language? For Arabic and Hebrew, confirm word spacing adjusts correctly under justification. | B | ☐ | ☐ | |
| 10 | Are orphan and widow controls enabled? Confirm the option preventing a single stranded word or line at a paragraph boundary is active. | B | ☐ | ☐ | |
| Category C · Fonts and typography | |||||
|---|---|---|---|---|---|
| No | Check and acceptance criterion | Grade | Self | Rev. | Notes |
| 11 | Is an appropriate font selected and embedded for each language? Verify Unicode block coverage, set PDF font embedding ON, and confirm zero corrupted glyphs. | A | ☐ | ☐ | |
| 12 | Does the font license cover the full multilingual distribution scope? Confirm the commercial font’s embedding, web, and print rights are sufficient for global distribution. | A | ☐ | ☐ | |
| 13 | Do font size, tracking, and leading meet per-language legibility standards? Apply the recommended leading for Chinese, Japanese, and Korean. Flag any body text below the 6 pt legibility floor. | B | ☐ | ☐ | |
| 14 | Do ligatures and combining characters render correctly? Inspect Arabic contextual forms, Devanagari conjuncts, and Vietnamese stacked diacritics. | A | ☐ | ☐ | |
| Category D · Visual elements | |||||
|---|---|---|---|---|---|
| No | Check and acceptance criterion | Grade | Self | Rev. | Notes |
| 15 | Do captions and numbering (Figure N, Table N) match the in-text references automatically? Apply automatic captioning and cross-references, then verify against the body text. | A | ☐ | ☐ | |
| 16 | Do leader lines and callouts avoid colliding with the translated text? Visually inspect whether expanded text pushes a leader line onto the artwork or clips it. | B | ☐ | ☐ | |
| 17 | Have table and chart cell widths and wrapping been adjusted for language expansion? Check for overflow inside cells, clipped header cells, and truncated chart legends. | B | ☐ | ☐ | |
| Category E · Multilingual behavior | |||||
|---|---|---|---|---|---|
| No | Check and acceptance criterion | Grade | Self | Rev. | Notes |
| 18 | Is the layout mirrored for RTL languages such as Arabic and Hebrew? Confirm page flow, running heads and feet, and table column order start from the right. Numerals and Latin characters stay LTR. | A | ☐ | ☐ | |
| 19 | Is CJK line-break prohibition (kinsoku) applied? Confirm the rules preventing prohibited characters at line start or line end are active. | B | ☐ | ☐ | |
| 20 | Do font and direction switch correctly in mixed-script runs? Check English model numbers inside Arabic body text, English abbreviations inside Japanese body text, and similar cases. | B | ☐ | ☐ | |
Findings summary
| Grade | Checks | Open | Closed | N/A | Sign-off rule |
|---|---|---|---|---|---|
| A Critical | 10 | Must be 0 to ship | |||
| B Major | 10 | Target 0. Log impact and fix plan |
Frequently asked questions
- How is this different from the translation integrity checklist? The subject is different. Translation integrity confirms that the translated file is fit to enter DTP. This checklist confirms that the laid-out deliverable renders correctly: page flow, text overflow, fonts, tables and charts, RTL and CJK behavior. The acceptance basis differs too. Translation integrity is judged on the translation file. DTP is judged on the final PDF or HTML.
- If even one Grade A item is open, can we still ship? The rule is zero before release. Grade A items are the ones that produce missing text, corrupted glyphs in the PDF, a broken generated TOC, or a reversed RTL table, all of which force a full regeneration or a reprint. If you override the rule under deadline pressure, record the exception reason, the approver, the residual risk, and the remediation date in the sign-off record.
- Can the same person do both the self-check and the reviewer check? It is not advisable. A DTP operator is the least likely person to catch an error in a setting they configured themselves. Detection is highest when someone who has not seen the editing file reviews the output on its own terms.
- If the product has no RTL or CJK languages, can we drop those items? Mark them N/A and record the reason in Notes. Keep the items in the list rather than deleting them, since languages get added to programs later.
- Do we run all 20 checks in every language? How long does it take? Run the 10 Grade A items in every launch language. Grade B items can be sampled on low-risk releases. For a 60 to 120 page manual, a trained reviewer completes the self-check in roughly one to three hours per language, and the independent cross-check in about half that. Expanding languages and RTL languages sit at the top of that range.
Pre-Launch QA Series for Multilingual Manuals (8 parts)
The series follows a multilingual manual from translation intake to market release, one gate at a time. Each part pairs a guide with a self-check checklist.
- Part 1 · Translation integrity Catching defects in translated files before editing
- Part 2 · DTP Multilingual typesetting and layout integrity (this checklist)
- Part 3 · Regulatory Market-specific regulatory and standards marking
- Part 4 · UI matching Synchronizing product screens with the document
- Part 5 · Accessibility EAA, WCAG, PDF/UA, and Section 508 conformance
- Part 6 · Multi-channel output Consistency across PDF, HTML, and mobile
- Part 7 · Print Bleed, CMYK, font embedding, and print readiness
- Part 8 · Data management Turning source files and TM into reusable assets