Gives each speaker in your video a voice, made on your own computer with Piper. Pick from 158 voices in 49 languages; each one downloads the first time you use it, and no graphics card is needed.

What it does

A voice per speaker

Choose Piper as the Dubbing engine and pick a voice for each speaker. Generate all speeches dubs that speaker's lines: all of them, a range, or the selection.

Downloads a voice once

A voice downloads from Hugging Face the first time a line uses it. After that it works offline.

Stays loaded

Once a voice is loaded it stays in memory for the next lines, so only the first line with each voice waits for it.

Fits the subtitle

With Fit dub to subtitle duration on, each new clip is sped up or slowed down to last as long as its subtitle. The original clip is kept too.

Voices

Pick a voice in the language of your subtitles. The Voice list shows all 158, and each name gives the language, the region and the quality, as in Amy (English, United States, medium). Piper reads with the voice's own language and accent, whatever your project's language.

The quality is x_low, low, medium or high, and a higher one is a bigger download: x_low voices are 20 to 27 MB, high ones 60 to 130 MB. Most voices (109 of 158) are medium. A few come in several qualities, such as Lessac and Ryan in American English, which have low, medium and high.

Try a line with two or three voices before you dub a whole speaker. Every voice you try is downloaded and kept.

177 voices in 59 languages.

ar-jo 2
bg-bg 1
bn-bd 1
ca-es 3
cs-cz 3
cy-gb 2
da-dk 1
de-de 10
el-gr 3
en-gb 11
en-us 27
es-ar 1
es-es 5
es-mx 3
et-ee 1
eu-es 2
fa-ir 5
fi-fi 2
fr-fr 7
he-il 1
hi-in 3
hu-hu 3
hy-am 1
id-id 1
is-is 4
it-it 4
ka-ge 1
kk-kz 3
kn-in 1
ko-kr 1
ku 1
ku-tr 1
lb-lu 1
lv-lv 1
ml-in 2
mr-in 1
ne-np 3
nl-be 4
nl-nl 6
no-no 2
pl-pl 5
pt-br 4
pt-pt 1
ro-ro 1
ru-ru 4
si-lk 1
sk-sk 1
sl-si 1
sq-al 1
sr-rs 1
sv-se 3
sw-cd 1
ta-in 2
te-in 3
tr-tr 1
uk-ua 5
ur-pk 2
vi-vn 3
zh-cn 2

Where voices are kept

In the add-on's own models folder: one .onnx file per voice, with a small .onnx.json beside it.

System	Folder
Linux	`~/.local/share/subtitld/addons/piper-tts/models`
macOS	`~/Library/Application Support/subtitld/addons/piper-tts/models`
Windows	`%LOCALAPPDATA%\subtitld\subtitld\addons\piper-tts\models`

Updates delete the voices. Updating or uninstalling the add-on replaces this folder, so the voices you used download again the next time a line needs them.

Options

In Settings › Add-ons › Piper TTS (offline) › Configure.

Piper model directory

Default: empty

This version of the add-on doesn't read this setting. Voices always go to the add-on's own models folder (see Voices above).

Languages

Each voice speaks one language with one regional accent, so the codes carry a region: pt-br and pt-pt are separate voices.

ar-jo
bg-bg
bn-bd
ca-es
cs-cz
cy-gb
da-dk
de-de
el-gr
en-gb
en-us
es-ar
es-es
es-mx
et-ee
eu-es
fa-ir
fi-fi
fr-fr
he-il
hi-in
hu-hu
hy-am
id-id
is-is
it-it
ka-ge
kk-kz
kn-in
ko-kr
ku
ku-tr
lb-lu
lv-lv
ml-in
mr-in
ne-np
nl-be
nl-nl
no-no
pl-pl
pt-br
pt-pt
ro-ro
ru-ru
si-lk
sk-sk
sl-si
sq-al
sr-rs
sv-se
sw-cd
ta-in
te-in
tr-tr
uk-ua
ur-pk
vi-vn
zh-cn

Platforms

System	Download
Linux x86-64	93.3 MB
macOS (universal)	74.6 MB
Windows x86-64	70.4 MB

CI builds each download before a release, but doesn't run a test synthesis with it.

Troubleshooting

When a line fails, it's left without a dub. These are the reasons the add-on gives.

model_missingThe voice couldn't be downloaded, or Hugging Face doesn't have it. Check your connection and try again; an unfinished download is thrown away, so the next try starts clean.
internalThe voice wouldn't load, or the speech failed. Running out of disk space during a download ends up here too: free some space and try again.
bad_paramsThe subtitle has no text to speak.

License

Add-on

MIT

Engine

GPL-3.0-or-later

Piper (OHF-voice/piper1-gpl)

Voices

Per voice

Mostly CC-BY; some CC0 or custom

Before you use a voice's audio commercially, check its license in Piper's voice list: most are CC-BY, some CC0 or custom.

Piper (offline)

What it does

Voices

Where voices are kept

Options

Languages

Platforms

Troubleshooting

License

Versions

Other Voices add-ons