A Guide to NNSVS / ENUNU

With Voice Model Database. Maintained by xuu.


What is NNSVS?

NNSVS (Neural Network Singing Voice Synthesizer) is an open source voice synthesizer created by Ryuichi Yamamoto (r9y9). Thanks to its open source nature, anyone can contribute and create voice models for the synthesizer. Currently, NNSVS does not have an interface of its own, so the majority of users utilize ENUNU, created by CrazY.


What is ENUNU?

ENUNU (pronounced eh-nu-nu) is a Plugin for UTAU, created by CrazY. Through ENUNU, UTAU can be used as an interface for editing the score used when rendering output of NNSVS voice models.
There is also a version of ENUNU for OpenUtau, which is run in the background while using OpenUtau. By using ENUNU for OpenUtau, users can edit a range of parameters including Pitch, Dynamics, Tension, Breathiness, Voicing and Gender. It's still being actively worked on and is not always perfect, but allows for editing ENUNU vocals without the need of an external program such as VocalShifter or Melodyne.

NOTE: Downloading NNSVS is not necessary to run ENUNU, as ENUNU will automatically download it on initial start-up. For most users I would recommend just using ENUNU.

Tutorials

Most users are a little intimidated when getting into NNSVS / ENUNU, and that's totally understandable! It's a bit more technical than most voice synthesizers, but it's slowly being made more and more accessible. Below I'll be linking some tutorials on how to get everything set up, as well as on how to make your own NNSVS voice model.

NNSVS Voice Database Creation Tutorial by PixPrucer
Downloading and Setting up ENUNU in OpenUtau by haraao
秋山大智 AI / A.I.CHI - ENUNU Usage Tutorial by PixPrucer / Peeslubn (useful general resource!)
More tutorials coming soon...! It will take some time to compile resources properly.

Voice Models

Naturally, if you want to use NNSVS / ENUNU, you'll need a voice model! I'm currently maintaining a list of all publicly available voice models, and will try to keep up with upcoming ones too. Press the button to be taken to the list!

Publicly Available NNSVS / ENUNU Voice Models

Last Updated: October 23 2022

Some notes on this page - models are listed in alphabetical order, with as much data provided as possible. If I am missing data, please feel free to contact me with the correct information. I try to keep this list up to date but as the popularity of NNSVS grows I may fall behind at times. I cannot embed videos from Niconico, only YouTube. Do not ask me to download a model in order to access things like information or high quality illustrations, this is a voluntarily-run database and making things easily available should be standard!

Models by Language