Note

This document is produced by machine translation

I am not a native English speaker. Sorry if the expression is not clear enough

English | 中文

lol-vo-extract

Extracting League of Legends champion/skin voices and matching them to in-game events

Special thanks

• Most implementations are refactored from CDTB
  • Heartfelt thanks to this excellent open-source project!
• Thanks to the code from wwiser
  • Before discovering this project,I could only build parsers by analyzing/guessing binary patterns through reverse engineering
  • Its emergence allowed me to fix quite a few potential bugs in parsing bnk/wpk formats This further improved the program's event matching success rate

Notable on v2

After very LONG delays, and three straight weeks of squeezing time out of my regular job to refactor codes Finally I finished the vast majority of v2!

The program is WORKING now, I still haven't had the time to finish internationalizing the console outputs and docs!

In v2, the event matching success rate was greatly improved
But I think it is still necessary to keep the following ATTENTION. After all, I'm only analyzing data as a rookie unofficial data miner:

ATTENTION PLEASE:
lol-vo-extract does not guarantee 100% match for all events for all skin!

• Different skins have different production dates and packaging formats. The possibility of extracting unknown events still exists
• Sometimes the literal event and actual trigger are different!

Requirement

• node.js v24+ (best)
  • node.js v14.18.1+ (maybe still works)
• VGMStream (required if extract format is wav)
• Ravioli Game Tools (required if extract format is ogg)
• ZSTD (optional, required for unpack game)
  • It is strongly recommended that your node.js version be at least v22.15.0 or higher
    As built-in support for ZSTD since this version. This built-in support significantly improve data processing efficiency
    No longer need to specified the fileZSTD option (still can if specified, the program will prioritize using fileZSTD)

Usage

This program is a config-file-base node.js program. No command line. All behavior is specified and adjusted through the config files

The entry file is index.js. Audio files will be stored in @1voice, and dictation files will be stored in @2dict by default

cd lol-vo-extract
node index.js

Config Instructions

All configs support both .jsonc and .json file extensions.

config.runcom.json

Required
This config specifies which hero/skin voice lines should be extracted using which profile
Refer to the example in config.runcom.json.example

config.user.json

Required.
This config contains user-defined profiles and must be configured by the user
Refer to the example in config.user.json.example
For detailed definitions, see ExtractConfig in bases.zh-cn.d.ts (english version not yet)

Special Keys:

• $profile
  This key specifies the default profile to use when no profile is explicitly defined in config.runcom.json
• $base in a profile
  This key indicates which profile the current profile is based on, including configs from config.default.json

Profiles will be assigned/overridden based on their relationships before extraction

config.default.json

This file provides the most basic configs for some preset profiles
Users can completely override these configs in config.user.json

config.event-manual.json

This config contains internally events manually provided by the user,
to assist in cases where events cannot be extracted automatically