RTL Urdu Transcription: Bidirectional Text and Nastaliq Challenges
Urdu is written right-to-left in Nastaliq script. When Urdu speakers mix in English words or numbers, the text becomes bidirectional - flowing right-to-left overall but switching to left-to-right for embedded content. This creates display challenges that pure Devanagari or Latin transcription never faces. Understanding RTL mechanics helps you produce clean Urdu transcripts that render correctly across platforms.
By CapsAI · Updated 19 August 2026
Key takeaways
- Urdu text flows right-to-left; embedded English and numbers flow left-to-right within that frame.
- Nastaliq (نستعلیق) rendering requires specific font support - many platforms default to Naskh.
- Bidirectional control characters can accumulate invisibly and break display in subtitle files.
- Test Urdu transcripts on the actual publishing platform before finalizing.
Why RTL matters for transcription output
When AI generates Urdu text, the character sequence must include proper RTL markers for correct display. A sentence like 'میں نے Google پر search کیا' contains Urdu words (RTL) surrounding an English brand name (LTR). Without proper bidirectional embedding, 'Google' may appear at the wrong position in the sentence.
Subtitle files are particularly vulnerable because many players have limited bidirectional text support. An SRT file with Urdu content may display correctly in one player but appear jumbled in another.
Nastaliq vs Naskh rendering
Urdu is traditionally written in Nastaliq style - a calligraphic script where characters connect with a downward-flowing diagonal. Most digital platforms default to Naskh style (used for Arabic) which places characters on a horizontal baseline. Naskh is readable but feels wrong to Urdu readers, similar to reading English in Comic Sans.
For transcripts intended for Pakistani or North Indian Urdu audiences, Nastaliq rendering matters. For subtitle files, Naskh may be acceptable because it renders more reliably across platforms and font stacks.
Start with 3 free minutes
Put Accurate Captions Into Practice
Create editable subtitles with CapsAI, then review Indian names, brands and places using the workflow in this guide.
Handling numbers and English terms in Urdu transcripts
Numbers in Urdu can use Eastern Arabic numerals (۱۲۳) or Western Arabic numerals (123). Both are valid. AI transcription typically outputs Western numerals, which create bidirectional complexity within Urdu text. The number 2024 in a Urdu sentence requires an LTR embedding to display its digits in the correct order.
English brand names, technical terms, and abbreviations are common in Pakistani Urdu speech. Each insertion creates a bidirectional boundary. Transcription systems must insert appropriate Unicode control characters (LRE, RLE, PDF) or rely on the Unicode Bidirectional Algorithm's heuristics.
QA workflow for Urdu transcripts
Verify Urdu transcripts in three environments: a text editor with RTL support, the target subtitle player, and a mobile device. Check that English insertions appear at the correct logical position in sentences. Verify that numbers read in the expected order. Look for orphaned punctuation - Urdu full stops and commas can end up on the wrong side of a line.
CapsAI's Urdu RTL subtitle checker tool can detect common bidirectional issues in SRT files. Run it before importing into video editors that may not preview RTL correctly.
- Check English words appear at the correct sentence position.
- Verify numbers display in logical order (not reversed).
- Look for punctuation attached to the wrong word.
- Test in the actual publishing environment, not just a text editor.
Frequently asked questions
Does CapsAI output Urdu in Nastaliq or Naskh?
CapsAI outputs standard Unicode Urdu text. The visual style (Nastaliq vs Naskh) depends on the font used by your display platform. The text itself is script-agnostic.
Why does my Urdu subtitle look correct in the editor but wrong in the video?
Video players use different text engines. Some lack proper RTL support and display Urdu characters in reverse logical order. Test with the actual player before export.
Should I use Eastern or Western numerals in Urdu transcripts?
Match your audience's expectation. Pakistani content typically uses Western numerals (123). If your content targets literary Urdu audiences, Eastern numerals (۱۲۳) may be preferred.
Start with 3 free minutes
Create Your Next Subtitle with CapsAI
Upload a video and generate editable AI subtitles with 3 free minutes to get started.