Scripting Tools for Voice-Acted Colonists

This is a set of tools modders can use to help them build new voicepacks for the Voice-Acted Colonists Rimworld mod

Three tools are included:

Audio Normalizer - This normalizes the volume of your voice lines so they aren't too loud or too soft. It also converts them to the .ogg format that Rimworld supports, compresses them (while maintaining quality) and renames them to acceptible file names.

Def Generator - Rimworld mods require special Definition files. These tell the game what the mod changes. This tool automatically generates all of the Def files your voicepack needs to run.

Renamer/Converter (Experimental) - When working with voice lines, you will often have folders filled with hundreds of unlabeled audio files. It can be very tedious to listen to each one before sorting them into the approriate folder. This tool uses OpenAI Whisper to transcribe an entire folder and subfolders full of audio files, then rename each file to the voice line itself. So if a file named "xyz123.ogg" contains the line "They'll never see us coming!", this tool will rename that file to "They'll_never_see_us_coming!.ogg". Now you can quickly look over your list of files and move them to the correct folder. At the same time, it will convert the files to .ogg format to reduce file size without losing quality. It will also normalize the volume so it isn't too loud or too soft. Note that not all .ogg files are supported by Rimworld. This first converts to wav, then to Ogg Vorbis. This ensures the format is correct. Currently this tool works perfectly for me, but I will need feedback to ensure it works well for you.

How to use the Audio Normalizer

1. Extract the NormalizeAudio folder

2. Your file structure MUST be the same as this voicepack

3. Sounds/VAC/YourProjectName is the project folder. Name your project folder whatever you like.

4. Put your voices in the Male, Female, or Any folders

5. Put your sound files in Attack, Select, Move, Downed, and Death folders

6. Double click "NormalizeAudio.bat". This automatically converts, normalizes, and renames all audio files for your mod.

7. Make sure you delete the .bat, .ps1, and .log files before uploading your voicepack to Steam

How to use the Def Generator

1. Extract the GenerateDefs folder

2. Your file structure MUST be the same as this voicepack

3. Sounds/VAC/YourProjectName is the project folder. Name your project folder whatever you like.

4. Put your voices in the Male and Female folders

5. Put your sound files in Attack, Select, Move, Downed, and Death folders

6. Double click "GenerateDefs.bat". This automatically generates all of the SoundDefs and VoicePackDefs for your mod.

7. Make sure you delete the .bat and .ps1 files before uploading your voicepack to Steam

Here's how it works (and doesn't work):

It scans for all sound files in the Sounds/VAC/Project/Gender/Voice/Action folders. It adds appropriate defs for each of these into the SoundDef and VoicePackDef for that voice. If a folder exists that is empty, it does not add a def.

Male, Female, and Any gender folders are now supported.

Currently only .ogg files are supported, and only Attack, Select, Move, Downed, and Death actions are supported. All of these can be expanded easily at a later date.

These scripts have been verified to work fine with this voicepack and with the DirtyBomb voicepack, but have not been tested with any other voicepack. As long as your file types and folder structure match, it should work fine. If you have any issues, you can bring it up on this repo, in discord, or you can fork this and modify to suit your needs.

How to use the Renamer

1. Download a specific version of Whisper: https://github.com/Purfview/whisper-standalone-win. Be aware that this download is over 1GB compressed, and almost 4GB when unzipped

2. Extract the downloaded Faster-Whisper-XXL

3. Place the contents of the "RenameAudio" folder in the folder which contains Faster-Whisper-XXL.exe

4. Copy your voice line audio files into the RenameThese folder

5. Make sure you close any audio players or anything that might be using these files

6. Run "RenameAudio.bat"

All audio files in the RenameThese folder and any subfolders will be renamed and converted

Audio files that are shorter than one or two seconds, as well as some files that only contain grunts but no actual words, will be renamed with a number (1.ogg, 2.ogg, etc)

If you have an Nvidia GPU, Whisper will use it to process your files very quickly. If you have an AMD GPU, your CPU will be used instead. With my CPU (i9 12900K), this will process about 70 voice lines per minute.

Question marks are not allowed in Windows file names, and will be replaced with a "~" symbol.

Currently the script shows only a small indicator that it is processing. Later I will add a more complete progress indicator.

The Transcription folders will be left for debugging purposes.

If you run into any problems, send me both of the log files that are generated.