ROM Hack Translation Xenoblade 2 - Delocalization & Retranslation Patch

  • Thread starter Thread starter rsc-pl
  • Start date Start date
  • Views Views 57,142
  • Replies Replies 173
  • Likes Likes 39

rsc-pl

Well-Known Member
Member
Joined
Jul 9, 2014
Messages
388
Reaction score
424
Trophies
2
XP
2,147
Country
Poland
Disclaimer​
If you don't intend to use this patch and only want to comment for trolling purposes, don't bother. The world doesn't need any more clowns. You're not special.

Xenoblade-2-Retranslation-Unburger-Mod-001.jpg


About the Patch

This patch brings the English version of Xenoblade 2 much closer to the original Japanese game. It restores names and terminology, corrects mistranslated or rewritten dialogue, preserves honorifics and cultural references, and removes localization changes that altered characterization, lore, jokes, terminology or the meaning of the original text.

The work now covers the main story, Iira, regular and mini quests, Blade Quests, Bonding Talks, inn scenes, field and NPC dialogue, tutorials, battle and system text, character and NPC names, enemy names, locations, weapons, attacks, skills, items and many other parts of the game.

Changes can be followed through the project's GitHub commits and release changelogs.

Xenoblade-2-Retranslation-Unburger-Mod-005.jpg


Core Philosophy

The purpose of this patch is to follow the original Japanese version of Xenoblade 2, not to preserve the Western localization where it conflicts with it.

Names, terminology, characterization, cultural references, jokes, religious material and story details are restored or corrected whenever the localization changed, replaced, simplified or removed them. The patch does not try to make the game more "Western", more "accessible", or more familiar to players of the official English release. That already exists. This project exists specifically to provide the opposite. An English version that respects what was actually written in Japanese.

Natural English still matters. The goal is not stiff word-for-word translation or Japanese sentence structure copied directly into English. The meaning, tone, characterization and concrete information of the Japanese text come first; the English is then written naturally without rewriting the source into something else.

Japanese cultural elements are not removed merely because they may be unfamiliar to a Western player. Honorifics are preserved. Genuine foreign-language names and words used by the Japanese script are preserved where appropriate.

This philosophy is not up for debate. Feedback about mistakes, missed details, technical problems, better readings of the Japanese text or places where the patch itself is inconsistent is welcome. Requests to restore localization rewrites simply because they are familiar from the official English release are outside the scope of the project.

Xenoblade-2-Retranslation-Unburger-Mod-002.jpg



What the Patch Changes

Dialogue, Story and Side Content
  • - The entire game, including Iira: The Golden Country, has been retranslated from the original Japanese script.
  • - This includes the complete main story, voiced and unvoiced scenes, regular quests, mini quests, Blade Quests, Bonding Talks, inn scenes, field dialogue, NPC dialogue, tutorials, DLC content and other text throughout the game.
  • - The translation follows the meaning, characterization and intent of the Japanese script rather than the rewritten dialogue of the Western localization. Information removed from the official English version has been restored, while additions and rewritten characterization that were not present in Japanese have been removed.
  • - The English has been written to read naturally rather than mechanically copying Japanese sentence structure.
  • - Japanese forms of address and honorifics are preserved throughout the game, together with differences in how individual characters speak to and refer to one another.

Names, Attacks, Enemies and Terminology

  • The patch uses the original names from the Japanese script throughout the game, replacing names and designations invented or altered for the Western localization.
  • This includes characters, NPCs, enemies and monsters, regions, cities, landmarks, facilities, organizations, quests, weapons, items, Arts, attacks, Blade and Driver skills, and other named elements throughout the main game and Iira: The Golden Country.
  • Japanese personal names follow their original order, and romanization is kept consistent throughout the game, including long vowels where appropriate, such as Menou rather than "Meno".
  • Names and terms that are deliberately written in another language in the Japanese version retain that identity instead of being replaced with unrelated English localization names. This includes English, German, Italian, French and other foreign-language elements used throughout the game.
  • The World Tree floors use their original Talmudic names — Vilon, Rakia, Shehaqim and Zebul — rather than the replacement names created for the Western localization.
  • Hana's transformation names are restored to their original forms, including Judgement Day (JD) instead of "Quantum Technochampion".
  • Core terminology follows the Japanese game consistently across dialogue, menus and gameplay text. This includes Divine Beast instead of "Titan", Eden instead of "Elysium", Bonding Talk instead of "Heart-to-Heart", Man-Eater instead of "Flesh Eater", Down/Rising instead of "Topple/Launch", and Pope/Holy See instead of "Praetor/Praetorium".
Battle, System Text and Tutorials

  • Battle and system text is translated from the Japanese rather than retaining the rewritten terminology and descriptions of the Western localization.
  • Arts, attacks, Blade and Driver skills, status effects, buffs, battle mechanics, equipment effects and other gameplay descriptions use the terminology and meaning of the original game.
  • Mechanical descriptions preserve the actual conditions and effects stated in Japanese, including activation requirements, damage conditions, Recast effects, reactions, elemental orbs, Party Gauge usage and other battle systems.
  • The game's tutorials have been retranslated from the Japanese, covering battle mechanics, progression systems, Blade mechanics, Chain Attacks, Specials, Driver Combos, Challenge Battle and the other systems explained throughout the game.
  • Terminology remains consistent between tutorials, menus, battle text, skill descriptions and other parts of the interface instead of using different names for the same mechanic in different places.
  • Where the interface has limited space, text is adjusted to fit the game's actual UI constraints while preserving the same meaning and terminology used elsewhere.
Nopon Dialogue and Characterization

Nopon dialogue follows the way Nopon actually speak in the Japanese script rather than the speech style invented for the Western localization.

Localization-only constructions such as Grampypon, Unclepon, biggipon/littlepon, brotherpon/sisterpon and similar invented -pon forms are not used unless the Japanese itself calls for such a form. The localization's forced broken English, automatic third-person self-reference and habit of inserting generic "friends" into Nopon dialogue are likewise absent where they have no counterpart in the original text.

At the same time, the patch does not flatten Nopon speech into ordinary English. Their actual speech patterns, characteristic sentence endings, verbal quirks, sounds and individual ways of speaking are preserved wherever they occur in Japanese.

The same principle applies to characterization throughout the game: characters retain their original forms of address, honorifics, recurring expressions, nicknames and differences in register instead of having those distinctions rewritten or standardized for the Western localization.

Xenoblade-2-Retranslation-Unburger-Mod-007.jpg


Important Restored Terminology

- 巨神獣 → Divine Beast, rather than "Titan". 巨神獣 literally refers to a giant divine beast, and the game itself connects them directly to the theological structure of the setting.
- 楽園 → Eden, used consistently rather than the localization's Elysium/Paradise mixture.
- Bonding Talk, rather than "Heart-to-Heart".
- Torigo, rather than "Trigo".
- Pope / Holy See, rather than Praetor / Praetorium.
- Man-Eater (マンエーター), rather than Flesh Eater.
- Down / Rising, following the terminology actually used by the Japanese game.
- Assist Core and Blade Resonance, Core Baptism used consistently across the game.
- Nia's original characterization of Zeke, including the turtle-themed nicknames and wordplay, has been restored.
- Original Japanese honorifics such as -san, -sama, -chan and -kun are preserved where the Japanese uses them. In Xenoblade 2 they frequently carry characterization and are part of scenes built around Japanese and otaku culture, so removing them changes more than simple politeness level.

Xenoblade-2-Retranslation-Unburger-Mod-006.jpg


Project Origins

I first experienced Xenoblade 2 through the Japanese version shortly after the Western release, so my understanding of its characters, story and lore came from the original Japanese presentation rather than the localization. Years later, revisiting the game became an opportunity to build on earlier delocalization work and push the restoration much further.

The technical side required reliable tools for Monolith Soft's BDAT format, along with a workflow for comparing, searching and editing thousands of game records. At this point the project represents several hundred hours of work, including translation, research, consistency checks, tooling, testing and repeated revisions across the game.


Prior Work

This project builds on groundwork laid by earlier community projects:
- Name Delocalization Mod
- Heart-to-Hearts Uncensored Mod


Technical Notes

The patch works through normal Switch romfs mod folders. It uses BDAT files from the Japanese version that are not present in the Western release, allowing the Western version to run with the Japanese text-data structure while using the translated patch content.

This also means:
- The game displays the original Xenoblade 2 logo in menus.
- The control guide menu remains in Japanese. This does not affect normal gameplay.

The game has many different interface and text-box limits. Text has been adjusted where necessary to avoid truncation while preserving meaning; these limits are not treated as one universal character count because different UI surfaces have different available space.

Xenoblade-2-Retranslation-Unburger-Mod-008.jpg



The Furigana / Ruby Text System

The Japanese game uses ruby/furigana-style text for more than simple pronunciation. In some places it creates a second layer of meaning by displaying one word while instructing the player to read it as another.

For example:

Code:
[System:Ruby rt=デバイス]僕[/System:Ruby]

The visible text uses 僕, while the ruby reading is "Device", adding information about how the speaker understands his own role.

Another important example is:

Code:
[System:Ruby rt=とうさま]神[/System:Ruby]

The displayed word is 神 (God), while the intended reading is とうさま (father). The choice of とうさま is itself meaningful characterization, especially when contrasted with another character using the much rougher おやじ for the same person.

The patch preserves this system because these layered readings can carry lore, characterization and cultural information that disappears if the ruby layer is discarded.


Installation

Install the patch like any other Switch romfs mod.

It is intended for the US version of the game with update 2.1 and all DLC/add-ons installed. It has been used on both emulation and Switch with CFW and is compatible with graphics mods.

When you start a new game, you will see a message in Japanese asking you to choose a language. Select Japanese (日本語): press “X” and then confirm with “A”.

IMPORTANT — FINAL CUTSCENE VIDEO ADD-ON
The final cutscene and credits use video files that are missing from the English release. Without them, the game can crash during the ending. The installation is therefore split into the regular patch and a one-time video add-on of roughly 400 MB. The video add-on only needs to be installed once.



This is how the mod directory should look.

Color legend:
- Files included in the one-time video add-on
- Files included in the regular patch

Emulator:


Emulator/​
├─ load/​
│ ├─ 0100E95004038000/​
│ │ ├─ Xenoblade 2 Unburger/​
│ │ │ ├─ romfs/​
│ │ │ │ ├─ bdat/​
│ │ │ │ │ ├─ jp/​
│ │ │ │ │ │ ├─ XXXX.bdat​
│ │ │ │ │ ├─ jp.bin​
│ │ │ │ ├─ stream/​
│ │ │ │ │ ├─ movie/​
│ │ │ │ │ │ ├─ mv10480100_h.webm​
│ │ │ │ │ │ ├─ mv10480100_s.webm​


Switch / CFW:

atmosphere/​
├─ contents/​
│ ├─ 0100E95004038000/​
│ │ ├─ romfs/​
│ │ │ ├─ bdat/​
│ │ │ │ ├─ jp/​
│ │ │ │ │ ├─ XXXX.bdat​
│ │ │ │ ├─ jp.bin​
│ │ │ ├─ stream/​
│ │ │ │ ├─ movie/​
│ │ │ │ │ ├─ mv10480100_h.webm​
│ │ │ │ │ ├─ mv10480100_s.webm​


Translation Comparison Examples

A few examples of how named scenes differ between the Japanese script, the official localization and the patch:

Original JapaneseOfficial LocalizationPatch
ヴァンダムが遺したものRex Makes His Mind UpWhat Vandham Left Behind
スザク誕生Suzaku Talks Kind Of Like VandhamBirth of Suzaku
黒鉄の城Birth Of A Marvelous New BladeThe Black Iron Castle
聖杯大戦Adel Spars with LauraThe Holy Grail War
雷轟のジークBringer of Chaos, Round ThreeZeke of the Thundering Roar
難民キャンプNo More Blades, No More WarThe Refugee Camp
雨I Lived As You WantedRain
愚かなるものたちThe Birth of a UniverseThe Foolish Ones



Companion Xenoblade Wiki Userscript

An optional Tampermonkey / Violentmonkey userscript is available as a separate download. It replaces localized character names, location names, terminology and other text across many of the most commonly used Xenoblade wikis so they match the patch.

This makes those wikis much easier to use while playing, without constantly having to translate localized names back into the names used by the patch.


Development Tools and GitHub

Two Python tools were developed for the project:
1. A BDAT extraction/repacking tool for managing game files.
2. A translation assistant that displays Japanese and translated text side by side and supports searching and editing the game's script.

- BDAT Tool
- Translation Assistant
- Patch GitHub


Contributing and Bug Reports

Useful contributions include:
- Bug reports and crashes.
- Typos, missing punctuation and text-overflow problems.
- Translation errors or missing information compared with the Japanese script.
- Localized names or terminology that were overlooked.
- Japanese-language research relevant to names, lore, characterization or cultural references.
- General testing and feedback from people actually using the patch.

When reporting a translation issue, Japanese-source evidence is especially useful. The official English localization is not treated as the authority for translation decisions in this project.


Latest Release - 0.9.0

Version 0.9.0 is one of the largest updates to the patch so far. It includes further corrections throughout the main story and side content, a large cleanup of remaining localized NPC/enemy/weapon/attack/skill/location names, revisions to battle and system text, corrected tutorials and gameplay descriptions, further Nopon-dialogue cleanup, and another broad consistency pass across terminology and UI text.

A separate full 0.9.0 changelog is provided with the release for anyone who wants the detailed list of notable name and terminology changes.

Older per-version forum changelogs have been removed from this post because they no longer reflect the current state of the patch. The GitHub history remains available for anyone who wants to trace earlier development in detail.
S6.jpg

S7.jpg
 

Attachments

Last edited by rsc-pl,
Extremely excited to try this out!
I would definitely suggest setting up a detailed progress overview (including listing things you have as TODO) so that we can know where to expect changes (and potentially find glitches or mistakes), and conversely knowing where to look for things that need changing. Similarly, a list of common changes would be helpful for identifying puns and such that have been overlooked.
I'm also curious about your methodology/scope: Is your plan to start with a coarse pass to cover specific significant reoccurring changes, and then eventually do a finer line by line pass? Or is the scope just covering those issues you've identified, and leaving the existing translation as is otherwise?
Also, I'm a bit confused with the stuff about restoring furigana - as far as I'm aware, the japanese version doesn't have any options to display furigana. Was there a system in place that you were able to enable that just was never exposed in the retail version for some reason? Screenshots of this would definitely be helpful.
 
Last edited by Niko_of_Death,
  • Like
Reactions: rsc-pl
Extremely excited to try this out!
I would definitely suggest setting up a detailed progress overview (including listing things you have as TODO) so that we can know where to expect changes (and potentially find glitches or mistakes), and conversely knowing where to look for things that need changing. Similarly, a list of common changes would be helpful for identifying puns and such that have been overlooked.
I'm also curious about your methodology/scope: Is your plan to start with a coarse pass to cover specific significant reoccurring changes, and then eventually do a finer line by line pass? Or is the scope just covering those issues you've identified, and leaving the existing translation as is otherwise?
Also, I'm a bit confused with the stuff about restoring furigana - as far as I'm aware, the japanese version doesn't have any options to display furigana. Was there a system in place that you were able to enable that just was never exposed in the retail version for some reason? Screenshots of this would definitely be helpful.
Thanks for your comment.
Regarding how the translation will look - initially I wanted to translate everything from scratch, but upon closer inspection of the game files, the scale of everything quickly brought me back to earth - the amount of dialogue and the scale of everything in this game is so enormous that it's practically impossible for a single person to finalize as a hobby project.
That's why I'll focus more on changing text that critically needs modification, along with adding many things suggested in the comments in places where I posted info (which will certainly appear).
Many characters are mischaracterized. The main plot has a quite different tone due to the lore itself (this has mostly been fixed already, but there are probably still many references to religious symbolism and folklore that need to be restored).
I want to change most of Hana's lines, for example. In many places, it's not that the words themselves are wrong in the translation, but primarily the tone of speech. Hana is really a different character than the one presented in the localization.
Localizers decided to make this character very childish. And that's not really how it is in the original.
Each successive form of Hana represents her developmental stage. JS, JK, JD, besides being developed by Tora as Jet Spark, Justice Knight and Judgement Day, are terms for girls in respective years of education - elementary school, high school, and university, which is directly reflected in the appearance of her successive versions and also in her speech style. In the Western version, these are just cringy puns.

Regarding progress overview - I will definitely add what has been fixed or changed, but it won't be anything very detailed like documents with every changed line or word - there's too much of it and maintaining such documentation, while looking professional and allowing everyone to see the scale of changes, simply takes too much time.
What I can do is write a script that will extract all the changed lines from the files and compare them with the original and this patch, then export them to a CSV/JSON file that can be imported into Excel or presented as a website.

As for furigana readings - of course the original version has them (A LOT OF THEM - when you search through Japanese files, you'll find over 4000 of them - many are repeated though). That's one of the fundamentals of building the narrative.
Clipboard01.jpg
 
Last edited by rsc-pl,
Ah, so its just applied by default in limited cases, not for all kanji like, say, dragon quest does. What are the technical constraints? I assume for proper centering youd need to manually check the alignment? (e.g. To have "fire" display above "homura" youd do "ho" "m"[fire] "ura")
 
Ah, so its just applied by default in limited cases, not for all kanji like, say, dragon quest does. What are the technical constraints? I assume for proper centering youd need to manually check the alignment? (e.g. To have "fire" display above "homura" youd do "ho" "m"[fire] "ura")
A certain number of characters. Oh, and you can't use spaces, but any "empty" ascii character works. This will always be displayed in the place where it is inserted.
It works like this:
Code:
[System:Ruby rt=Ars - Art or craft in Latin ]Divine Beasts[/System:Ruby]

Ars - art of craft in Latin
  Divine Beast
 
Hi, you said that its tested on Citron but im having troubles getting it to run, could you tell me where exactly you placed the mod folder?
 
Hi, you said that its tested on Citron but im having troubles getting it to run, could you tell me where exactly you placed the mod folder?
You have to start the game, then put them in the load folder. Its the same as it was for Yuzu
 
You have to start the game, then put them in the load folder. Its the same as it was for Yuzu
Well I did that, I have other mods installed and running on the game (they show up in the add-ons tab) but this one just doesn't seem to work? I'm at a bit of a loss as the only difference between this one and the rest is that it lacks an exefs folder.
Post automatically merged:

Well I did that, I have other mods installed and running on the game (they show up in the add-ons tab) but this one just doesn't seem to work? I'm at a bit of a loss as the only difference between this one and the rest is that it lacks an exefs folder.
nvm files were too nested... thanks for the help :)
 
Last edited by Hehexdaylmao,
  • Like
Reactions: rsc-pl
It is not a foreign name, so it is a completely separate case from Arisu. Rinne is a real Japanese name.
Just because a name is on that site doesn't mean its a traditional japanese name, Arisu is also listed there https://japanese-names.info/first-name/arisu-2/


I asked my coworker (who is native japanese), and she said she'd associate it with the english name (of course, shes not a language expert so thats anecdotal).

If it were an explicitly japanese character it would make some sense, but given her appearance/character I'd say an English name was the intent.
 
  • Like
Reactions: Legitendo
Man, I wish someone would do the same for Breath of the Wild (and maybe Tears of the Kingdom, but I haven't played it yet). I hate hearing something in Japanese and reading something else in English.
Thanks for your work btw
 
Last edited by SkullHex2,
  • Like
Reactions: rsc-pl
Disclaimer​
If you don't intend to use this patch and only want to comment for trolling purposes, don't bother. The world doesn't need any more clowns. You're not special.


This project represents a passionate effort to restore Xenoblade 2 to its original vision, building upon previous delocalization work while introducing substantial improvements. While previous mods laid important groundwork, this patch goes further in preserving the game's authentic Japanese cultural elements and narrative nuances that were altered or removed in the Western release. In addition, many things that previous patches left out by accident have already been fixed (not all names and designations have been restored, leaving some localized versions elsewhere in the game).

At the outset, I would like to mention that the patch is work in progress (and will remain so for a long time, unless I find a lot of people willing to help). Many things may not work as intended, some dialogues will need editing after initial translation, etc. You use at your own risk. However, any report of any bug, or providing something I have overlooked that has a significant impact on the perception of characters, story or lore will be greatly appreciated. Changes can be checked in the github's commits.

Prior Work

- Name Delocalization Mod
- Heart-to-Hearts Uncensored Mod


A Personal Note on Project Origins

My journey with Xenoblade 2 began with the Japanese version, which I completed shortly after the Western release. This means my understanding of the characters, story, and lore comes purely from the original Japanese presentation, uninfluenced by the Western localization. Recently, I felt drawn to revisit this masterpiece and realized it was the perfect opportunity to refresh existing delocalization projects.

However, this wasn't a straightforward task. The main technical hurdle was working with Monolith Soft's proprietary BDAT file format. It wasn't until recently that I discovered a reliable method for unpacking and repacking these files. The current workflow involves processing JSON files through Notepad++ and a custom tool I developed specifically for this project that facilitates file searching, editing, and translation work.

As of the date of writing this post (17.02.25), I have already put well over 30 hours into this patch.

Core Philosophy

The fundamental goal of this patch is authenticity to the original Japanese version without compromising cultural elements for the sake of "Western accessibility." This isn't about making the game more "digestible" for the average player - the official localization already serves that purpose even removing the entire original lore and replacing it with “more appropriate” counterparts. Instead, this patch aims to preserve the rich cultural context and nuanced storytelling of the original, particularly its roots in Christian symbolism, Japanese and Chinese folklore and connections to otaku culture everyhere which were all significantly downplayed (or rather completely removed) in the Western release.

Major Changes and Restorations
(in addition to changes made in previous projects)


Terminology Unification and Cultural Accuracy


- Restored 巨神獣 to "Divine Beast" rather than the official "Titan" - this better reflects their divine nature and connection to the game's theological themes (in the original text specifying that they are しもべ/Servants of God). 巨神獣 Means "Giant Divine Beast" but it is generally too long to fit in many places in the game interface.
- Unified all instances of 楽園 as "Eden" (in previous patches inconsistent between "Paradise", "Eden", and "Elysium" was left from localization in many places). This choice was made specifically because the game's narrative centers around humans who once lived in a fertile land with God - directly paralleling the biblical Garden of Eden
- Restored "Bonding Talk" as the proper translation for the intimate character conversations (previously localized as "Heart-to-Heart")
- "Trigo" is now "Torigo"
- Restored all NPC names to their Japanese originals
- Completely revised and retranslated all attack names, restoring their original English, German and Italian names that were completely replaced by localizers
- Restored the World Tree's floor names to their original Talmudic references (names of Heavnes) - Vilon, Rakia, Shehaqim, and Zebul - rather than the Western localization's star-based naming system.
- Implemented proper waapuro romaji for character names, precisely - respecting long vowels (e.g., "Menou" instead of "Meno") - previous patches have mixed different methods, while I value consistency.
- Restored all instances of Hana's transformation names (e.g., "Judgement Day (JD)" instead of the localizer's "invention" - "Quantum Technochampion")
- Restored Nia's authentic characterization of Zeke, including her turtle-themed nicknames and related wordplay (such as turtlehaed)
- Added all original Japanese honorific suffixes (-san, -sama, -chan, -kun). This was a massive undertaking but rather essential for Xenoblade 2, as these suffixes are frequently used as characterization devices. The game makes numerous references to otaku culture - for example featuring scenes where Homura and Hana simulate Maid Cafe interactions - contexts where these honorifics are crucial to the scene's tone and impact. If the patch was about Xenoblade 1 or Xenoblade 3 - I would rather not undertake this approach, because there this aspect does not play such an important role to spend really a lot of time to search and restore every honorific in the dialogue.

Technical Breakthroughs

What distinguishes this patch from previous relocation attempts is the compatibility with normal mod folders (romfs) - something previous patches struggled with due to (probably) conflicts with game updates. Through investigation, I discovered a method to utilize bdat files from the Japanese version that don't exist in the Western release, thus avoiding update conflicts entirely. Although the Western version does not have the Japanese language, after copying the files, the Western version runs in the original mode. This has two effects:
- The game now displays the superior original Xenoblade 2 logo in menus
- The control guide menu screen remains in Japanese, but this is the only affected screen and shouldn't impact gameplay at all

The Furigana Text System Restoration

One of the omissions in the Western release was the abandoning of the Furigana Text System which could have served in the Western version not only as a counterpart to what it was used for in the Japanese version, but also as a clarification of the meaning of the names, which would have resulted in the fact that none of the names “had” to be changed (the meaning of the names is the only argument of the proponents of changing them). The original game uses this system for additional context and depth through tags like:

Code:
[System:Ruby rt=デバイス]僕[/System:Ruby]
This creates fascinating layers of meaning. For example, in the text above, while the character uses 僕 (boku, meaning "I/me"), it's meant to be read as "Device" - adding context about the character. An even more meaningful example is:

Code:
[System:Ruby rt=とうさま]神[/System:Ruby]
Here, while the text shows 神 (kami, meaning "God"), it's meant to be read as とうさま (tousama, meaning "father"). This reflects not only the profound connection with the story's themes, but also powerfully characterizes the speaker's relationship and attitude. The choice of とうさま (tousama) - a very respectful and formal way of saying "father" - contrasts sharply with another character who uses おやじ (oyaji) - a much more casual, almost rough way of saying "dad" or "old man" - for the same person. These different word choices for "father" reveal volumes about each character's personality and their relationship with their father figure, adding layers of characterization that were simply lost in the Western release.

This system can and (will) serve multiple crucial functions in this patch:
- Provides essential context about the world, characters, and lore
- Enables the inclusion of short translation notes explaining character names and their origins
- Offers cultural context that enhances understanding of Japanese cultural elements
- Works seamlessly in both dialogue subtitles and NPC speech bubbles
- Completely nullifies the "localization argument" about changing character names, as we can now explain their meaning directly in-game

Technical Constraints

The engine has specific limitations that required careful attention:
- Strict 56-character limit per line
- Maximum of two lines per dialogue box (112 characters total)
- Text exceeding these limits is truncated
Previous mods sometimes overlooked these constraints, resulting in cut-off text. Every line in this patch has been carefully adjusted to respect these limits while preserving meaning.

Development Philosophy

This patch follows a rolling release model - rather than waiting for "completion," new versions will be released periodically for testing. Community feedback on bugs, translation errors, and overlooked details is invaluable and will help drive continuous improvement and restoring this game to be as good its original Japanese release.

However, let me be absolutely clear: while the feedback is welcome, the fundamental translation philosophy is not up for debate. This project exists specifically to provide an alternative to the heavily localized Western release, focusing on authenticity to the original Japanese version.

Installation
Just install it like any other switch romfs mod.
It's tested on both - emulation (Citron) and Switch with CFW. US version of the game with 2.1 patch and all addons installed. It's compatible with graphics mods too.


Screenshoots

View attachment 486606
View attachment 486607
View attachment 486608
View attachment 486609
View attachment 486610
View attachment 486611
View attachment 486612


Localization Comparison Examples

To illustrate just how the Western localization departed from the original text, here are some striking examples of chapter names:

Original Japanese | Official Localization | Translation
-----------------|----------------------|---------------------
ヴァンダムが遺したもの | Rex Makes His Mind Up | What Vandham Left Behind
スザク誕生 | Suzaku Talks Kind Of Like Vandham | Birth of Suzaku
黒鉄の城 | Birth Of A Marvelous New Blade | The Black Iron Castle
聖杯大戦 | Adel Spars with Laura | The Holy Grail War
雷轟のジーク | Bringer of Chaos, Round Three | Zeke of the Thundering Roar
難民キャンプ | No More Blades, No More War | The Refugee Camp
雨 | I Lived As You Wanted | Rain
愚かなるものたち | The Birth of a Universe | The Foolish Ones

Development Tools

To facilitate this project, I developed two Python programs:
1. A BDAT file repacker for managing game files
2. A dedicated translation tool that displays Japanese and translation scripts side by side and allows to edit the translation.

GitHub Repositories:
- BDAT Tool
- Translation Assistant
- Delocalization & Retranslation Patch Github

Contributing

This project welcomes contributions in several areas:
- Translation (for those who share the same philosophy of authenticity)
- Testing and bug reporting
- Original Japanese lore consultation (not based on localized versions)
- General feedback and suggestions

This isn't just another localization - it's an effort to restore Xenoblade 2 to its authentic form.

If there are more interested people, I will create a group on discord.

News
*25.02.2025*
v0.0.24 Dialogue of the main plot revised up to the end of the third chapter.
Man-Eater (マンエーター) is used in my patch instead of Flesh Eater.

*23.02.2025*
Added v0.0.21 0.0.23 Many fixes, main story revised up to the middle of the 3rd chapter.

*20.02.2025*
All names and proper names are now consistent with the original. Dialogue of the main plot revised up to the end of the second chapter.

*18.02.2025*
First release.
huh this is neat, never knew they botched localization, I always thought that the voice acting sounded weird was intentional compare to the japanese dialogues
 
huh this is neat, never knew they botched localization, I always thought that the voice acting sounded weird was intentional compare to the japanese dialogues
yeah whoever did localization had a major hard on for The British isles and Greek for some reason, bit disappointing that as usual these people just bastardise everything in the name of "localization"
 

Site & Scene News