AI summary by TwinMind

Why don't subtitles match dubbing?

Tom Scott

8m 26s

Introduction and Core Problem

  • Video dubbed/subtitled into multiple languages
  • Subtitles and dubbing often don't match outside original English
  • Common issue in professional TV shows, frustrating for viewers who use both
  • Short reason: subtitles and dubbing done by different teams

Translation Team Approach Differences

  • NordVPN Ad Example Analysis
    • French, Spanish, Portuguese: word order similar to English, simple translation
    • Hindi: company name sits mid-sentence in obvious translation
      • Dubbing team: rephrased to put company name at end for lip sync matching
      • Subtitling team: kept standard translation, no lip sync constraints
    • Japanese: dubbing team found translation with close enough lip movements
  • Core Question: Why not use dubbing script for subtitles? Why translate twice?

Subtitle-Specific Limitations and Constraints

  • Reading Speed vs Speaking Speed
    • Modern approach: reading faster than talking, put every word on screen
    • Viewers can pause if needed
  • Historical Subtitle Practices (10-15 years ago)
    • Old subtitles often summarized rather than transcribed
    • Style guides emphasized "careful and sensitive subtitle editing"
    • Assumption: slow readers needed meaning over exact words
    • No easy pause/rewind capability then
    • Didn't anticipate mobile viewing or sound mixing issues

Complex Translation Example: Maroon 5 Rant

  • Original English: "Maroon 5's Memories is infuriating composition using Canon in D melody but never resolves, same stuck-in-head effect as Alfred Bester's The Demolished Man jingle, and I hate it"
  • Japanese Translation Decision
    • Removed Alfred Bester reference (unknown to Japanese audiences)
    • Conveyed same meaning in fewer characters
    • Better experience for subtitle readers vs breathless high-speed dubbing performance
  • Reality Show Challenge: 6 people shouting/interrupting
    • Subtitles must summarize
    • Voice actors can shout over each other in dubbing

Translation Philosophy: Words vs Meaning

  • Impossible Pun Example: "Anyone want to buy goslings? They're going cheep"
    • Requires knowledge: "going" = making noise, "cheep" = baby bird sound, "going cheap" = on sale
    • Translation Responses:
      • Some translated word-for-word, gave up on joke
      • Portuguese subtitle/dubbing teams: same joke using "ganso" (goose) + "cansado" (tired) = "gansado"
      • Spanish subtitler: same "gansado" solution
      • French/Spanish dubbing: created puns based on bird noise words in their languages
  • Animal Sounds Differ by Language
  • AI Translation Limitations (2023): Cannot capture nuance, meaning, performance - still requires humans

Technical Synchronization Challenges

  • Goose Honk Synchronization
    • Translators had to match goose's honk timing
    • Copied background noise from different video section
    • Had to "beak-sync" the goose (no lips)

Case Study: "Have It" Translation Challenge

  • Context: Roller coaster phobia video, moment of triumph on Nemesis ride
  • Original: "Have it!" - obscure British slang
  • Meaning: excitement, triumph, adrenaline, encouragement, success
  • Usage: after team scores goal, cheering friend after downing two pints
  • Character: very male, "drunken lads' night out" energy
  • Translation Challenge:
    • No literal meaning, difficult to Google
    • Regional variations in Britain
    • Translators given context notes, but most productions don't provide this
    • Only have original show, maybe script

Final Constraints Summary

  • Subtitling Teams: any word/phrase choice, must fit on screen
  • Dubbing Teams: voice actors can emote fully, but lip movements must match
  • Result: This is why subtitles and dubbing differ

Sponsor Integration

  • NordVPN Features: Quick Connect button, ~60 countries (Albania to Vietnam)
  • Use Cases: stop auto-translation while traveling, access geo-blocked content
  • Compatibility: Windows, Linux, Mac, iOS, Android (6 devices simultaneously)
  • Guarantee: 30-day money-back
  • Personal Endorsement: Actually uses product regularly, keeps being useful

TwinMind

Get summaries like this for your own meetings

TwinMind takes notes in the background during meetings, lectures, and calls, then writes the summary for you.

Free plan to get started

No bot joins your calls

Phone, desktop & Chrome apps

Your next meeting, summarized like this

TwinMind listens in the background and turns meetings, lectures, and calls into searchable notes you can chat with. Start free.

Available on phone, desktop & Chrome

Your next meeting, summarized like this

TwinMind listens in the background and turns meetings, lectures, and calls into searchable notes you can chat with. Start free.

Available on phone, desktop & Chrome

Your next meeting, summarized like this

TwinMind listens in the background and turns meetings, lectures, and calls into searchable notes you can chat with. Start free.

Available on phone, desktop & Chrome