Scylla's Band (offline, 14 voices, 6 languages)

What it does

Before you install

• Disk: about 300 MB, most of it the model, downloaded on the first line.
• Memory: under 500 MB while it speaks.
• Processor: any 64-bit Intel or AMD processor, or any Apple Silicon Mac.

Scylla's Band speaks only in its own voices: it does not clone the speaker in the video. For the speaker's own voice, use VoxCPM2, Sopro or Chatterbox.

Voices

In the Speakers tab, choose Scylla's Band as a speaker's Dubbing engine, then pick a Voice. Turn on Fit dub to subtitle duration for the speaker to stretch or squeeze each line to its subtitle. Set the project's language before you dub.

14 voices.

The fourteen voices are Scylla, Ariadne, Felix, Gwen, Ink, Kincade, Max, Orpheus, Poppy, Priya, Rex, Stone, Thistle and Tuesday. Each was trained in all six languages, so the list is the same whatever your project's language. In English, Ink, Orpheus and Tuesday speak British English and the others American English.

Subtitld gives each speaker a seed, so dubbing a line again gives the same take. Change a word and the line is spoken anew.

Models

The model is not inside the add-on: it downloads into Subtitld's shared model cache the first time you dub, from spybyscript's Hugging Face repository at a pinned release, each file checked. An interrupted download resumes where it stopped. Updating or removing the add-on keeps the model.

The add-on runs Scylla's Band's own C++ runtime with Google's LiteRT, inside the add-on's process, so it needs no Python machine-learning libraries and starts in a moment.

Options

In Settings › Add-ons › Scylla's Band (offline, 14 voices, 6 languages) › Configure.

Intensity, pace and pitch apply to every speaker who uses Scylla's Band in the project.

Languages

English, Spanish, Italian, French, German and Vietnamese. A project in any other language gets a message that says so: pick another dubbing engine for it.

Numbers are read as words in each of the six languages, so "42" becomes "forty-two" in English and "bốn mươi hai" in Vietnamese.

• en
• es
• it
• fr
• de
• vi

Platforms

Linux x86-64, macOS on Apple Silicon and Windows x86-64. Each release is built on all three, and on each one CI installs the add-on as Subtitld does, downloads the model and dubs with it in all six languages. On Linux it needs glibc 2.35 or newer.

Troubleshooting

• network_unavailableThe model couldn't be downloaded. Check your connection and try again: the download resumes.
• disk_fullNot enough disk space for the model, about 300 MB.
• unsupported_languageThe project's language is not one of the six, or not set. Set it in the project, or pick another dubbing engine.
• unsupported_voiceThe project names a voice this version doesn't have. Pick the voice again in the Speakers tab.
• bad_paramsThe line has no text.
• internalSomething failed inside the add-on; the message has the details.

License

The add-on, Scylla's Band's runtime, LiteRT and the Scylla's Band weights are all Apache-2.0, and the parts of the model that come from Vocos and DeepPhonemizer are MIT. You can use the add-on and the dubs it makes for any purpose, paid work included. The dubs you make are yours.