DisDubs › Dub packs
Guide
Dub packs
Updated 17 August 2026
A dub pack is a folder holding one scene, cut up so a group can perform
it. Inside: the video, the scene’s audio sliced into numbered clips of a few seconds
each, usually a dialogue-free _backing_track, and a metadata file that gives
every clip a timestamp, a caption and a character.
The format comes from The Choicer Voicer. There are hundreds of community packs, and they are the reason DisDubs reads that layout rather than inventing another one.
What is inside a pack
The layout is flat. One folder per scene, containing:
-
Numbered audio clips:
01_NiceOfThePrincess,02_LotsaSpaghetti, and so on. The numbering is what keeps them in order. The official guidance is to keep each one under about six seconds and cut at intuitive spots, which is a line of dialogue rather than a bar of time. -
dub_video.ogv: the picture, in Ogg Theora, because that is what the game’s engine supports. -
_backing_track: optional, and the thing that makes a dub sound like a dub. It is the scene with the dialogue stripped out: music, effects, room tone. It plays under the performances, so what you are replacing is only the voices. -
A metadata file: timestamps in seconds, captions, tags, portraits,
and the character each clip belongs to. In The Choicer Voicer you build it from
Extras → Edit Voice Packs & Clip Metadata; on disk it is an
.inior a.txt.
Clip audio is .ogg or .mp3. Filenames can contain spaces. Not
every pack has a _pack_info.ini, and plenty of good ones do not.
Where to find packs
- The official dub packs on itch.io, from the game’s developer.
- The Dub Mode category on GameBanana, which is where most community packs land, along with compilation packs bundling many scenes at once.
- Creators publishing packs directly on itch.io, usually themed around a show, a film, or a run of memes.
Whatever you download, you are responsible for having the right to use it. Most packs are cut from films and television nobody in this chain owns. Our terms and copyright policy say where that leaves everyone, and there is a takedown route that a real person reads.
Playing a pack with friends, over Discord
DisDubs is a Discord Activity that reads these packs. Upload the zip from inside Discord and the scene shows up in your server’s picker, ready to cast. Nobody installs anything, and it is the same on a phone as on a desktop.
What the importer does on the way in, so you do not have to:
-
Transcodes the video. Chromium dropped Theora, so
dub_video.ogvplays in no browser. It comes out as H.264/AAC. This is why an import takes a while after the upload finishes. The upload was never the slow part. - Pads the backing track. A bed shorter than the picture used to truncate the video and take every late line with it. Now only the video can end the scene.
- Unwraps a zip inside a zip, which is the ordinary mistake and not worth an error message.
-
Matches extensions case-insensitively. A pack with
.OGGfiles works on Windows and silently loses its audio on Linux; this is that bug, fixed. - Drops a line cued past the end of the picture and tells you, rather than clamping it or refusing the whole pack. A pack is only refused when every line is past the end, which means the timestamps are for a different video.
-
Reads the backing track’s meaning. A
_backing_trackis dialogue-free, so it plays at full level under your voices. A pack without one still has the original performance in it, so it is mixed down under the dub instead of over it.
DisDubs also reads DubForge packs, which name no speakers: the cue lives in
the clip filename, _TIMESTAMPS.txt is the fallback, and
_captions.json supplies captions.
The limits, stated plainly
A free server gets three upload slots and can import scenes up to 3 minutes long, at five imports every six hours per person. Studio raises the ceiling to 10 minutes, Studio Pro to 20. A pack file can be up to 2 GB, uploaded in chunks so a dropped connection costs one slice rather than the whole thing.
Uploaded scenes are private to the server that uploaded them unless you ask for them to be published. Nobody else’s server sees your packs.
Making your own pack
It is more editing than programming. The rough shape is the same whichever program you intend to play it in:
-
Pick a scene worth performing
Look for two to four characters, clear turn-taking, and a minute or two of runtime. Scenes where everyone talks over everyone are miserable to cut and worse to dub.
-
Split the audio from the video
Then, if you can, split the dialogue from everything else. A stem separator will get you a usable music-and-effects bed; save that one as
_backing_track. -
Cut the dialogue into clips
One line per clip, under about six seconds, cut where a person would naturally stop. Number them in order (
01_,02_,03_) and give each one a name you will recognise in a list. -
Write down when each line starts
Timestamps in seconds, against the video. This is the part that decides whether the finished dub lands on the mouth or a beat behind it, and it is worth being fussy about.
-
Add captions and characters
Captions give performers a script to read. Naming the character on each clip is what lets a group divide the scene up rather than everyone reading everything.
The Choicer Voicer’s official content guide covers its in-game metadata editor step by step, and a pack built to that guide imports into DisDubs unchanged.
Common questions
Where does the packs_voice folder live?
In The Choicer Voicer’s game files; the main menu will take you there. DisDubs has no folder to find, because packs are uploaded from inside Discord and stored against your server.
Do I have to convert the OGV myself?
No. Upload the pack as it is and the importer handles the video.
Can I upload a pack I did not make?
Technically yes, and it is a question of rights rather than of the software. Read the terms. Packs stay private to your server by default, which is the setting that most closely matches how people actually use them.
My pack imported but a line is missing.
That line was cued past the end of the video, so it was dropped and named in the import result rather than silently mangled. Usually it means the timestamps were written against a longer cut of the scene than the one in the pack.
What if the pack has no backing track?
It still works. Without a _backing_track the original dialogue is in the
scene audio, so it is mixed underneath your performances instead of replacing them, and
you can pull it down further with the bed slider before rendering.