NoLyui
DiffSinger Download


Terms and Conditions

1. Definitions
1.1 Voice Library
The “NoLyui” singing synthesizer model developed for DiffSinger by Kaishikuon and
voiced by IseulEasel, supplemented by data from Kaishikuon, CometP, and the つくよ
みちゃん UTAU Database Unofficial Research Corpus, henceforth referred to as “this
voice library”.
1.2 Voice Quality
The characteristics and tone of your voice. Broadly speaking, something that
“maintains your voice quality” is something that sounds like you and is recognizable
as your voice.
1.3 Assets
A work made using the voice library or character which is published for other people
to use to create their own works with
2. Commercialization
This voice library may be used for the following commercial purposes without
authorization:
• Original music (whether released on streaming services or sold directly for a
fee)
• Merchandise sales that do not feature the official artwork by IseulEasel
• Works that generate advertising revenue
• Sale of audio generated by the voice library directly, such as for a sample pack
or audiobook
• Commercialized non-musical works such as a video game or animation as an
uncharacterized voice
• Non-profit usage by a corporation
• Use in work commissioned by an individual such as hiring someone to create a
song or artwork
Please note that while credit is appreciated for all uses, it is required for use in works
that feature the voice but are not strictly musical works such as video games, sample
packs, audiobooks, video games, films, and animations. You must also provide credit
when publishing assets, whether paid or free, regardless of whether it uses the voice
or character.
The following commercial use requires permission from IseulEasel:• Use for a sponsorship or advertising with the voice library directly
• For-profit use by a corporation
• Use as a virtual “voice actor” or official voice for an unrelated character in a
commercial work
The following commercial use is prohibited under any circumstance:
• Sale of the voice library or derivative software created using the voice library
files or audio generated using it
• Sale of merchandise featuring the character’s official artwork
• Creation of blockchain assets
3. Usage Restrictions
Expression
This voice library may be used for work that is not appropriate for all audiences such
as sexuality and gore, as well as non-glorified extreme themes such as suicide,
addiction, abuse, murder, gun violence, child exploitation, or human trafficking. When
work is not suitable for all ages, you are required to add warnings and tag your work as
mature content.
The following expressions are prohibited:
• Discrimination against anyone on basis of race, ethnicity, sex, gender, religion,
social status, illness, or disability
• Attack of any person or group of people or to call for harassment
• Promotion of a cause, political party or politician, or ideology.
• Glorified extreme topics
• Broadly, anything that is widely accepted to be controversial
• Impersonation of IseulEasel or any use that implies the voice or character are a
real person
• Use with AI algorithms trained using data collected without the consent of the
original creators. For more detailed information on what kind of AI models are
permitted, please see the following section.
Modifications
You may modify audio generated using the voice library with filters and postprocessing techniques such as mixing and mastering plugins.
You may convert the audio into someone else’s voice quality, but you are required to
disclose that it is an artificial voice and provide credit. You may not convert the audio
into the voice quality of someone that did not give permission for you to do so, such as
a celebrity. Use with voice converters where the voice actor consented to the release
of the product count as express permission.
You may not alter the official artwork in any way that changes the original form such as
color edits, changes to proportions, tracing, or otherwise drawing on top of the image.
You may apply simple global filters like a red overlay or change the image to a
silhouette for works such as music videos. While you may change the size or image
format for the purpose of displaying the original image, you may not use any AI
algorithm that modifies the image. This applies both to generative AI that creates a
new image based off of the artwork and image quality AI such as upscalers or noise
removal algorithms.
You may not reverse-engineer the voice library files for the purposes of re-enabling the
voice quality of the supplementary data.
Characterization
When creating original works featuring NoLyui, you may make changes to official
character details such as clothing, hair and eye color, sex, gender, and age, but may
not change the character’s race*, ethnicity, or skin color. Slight changes to skin tone in
order to reflect lighting or artstyle are permitted, but you may not create the
impression of the character having a noticeably different skin tone.
*Restrictions on race only apply to races and ethnicities that exist in real life.
Depictions of NoLyui in a fantastical race such as an orc or elf are permitted, however,
you may not significantly change the skin tone even if it is to a color like blue or green
that does not exist in real life.
You may make your own design to represent the voice library and will keep copyright
ownership over the design should you do so. However, you may not make an
alternative virtual idol or “fanloid” that uses NoLyui (whether modified or unmodified)
as the character’s voice.
You may use NoLyui as the official voice for an unrelated fictional character in noncommercial works, such as a voice-claim for your original character, the voice for a
character in a song series, or as the voice of a character in a free video game.
However, using the voice library or derivative software to replace the voice of a real
person such as for a Vtuber requires prior permission from IseulEasel.
Derivative Software
You may use the voice library or audio generated using it for personal research such as
a private software that is not released to the public. You must provide credit if you post
audio generated publicly. You may not release any music-based or voice-conversion
software that maintains NoLyui’s voice quality for others to download, whether they
are an entirely new software or developed for an existing engine such as UTAU.
However, you may develop text-to-speech models.When developing software that generates audio of any kind, but does not maintain
NoLyui’s voice quality, you are required to provide credit. There are no credit
requirements for software that do not generate audio such as lip-syncing software.
Regardless of whether audio is generated or NoLyui’s voice quality is maintained, any
software created using this voice library must inherit all restrictions imposed by this
license. You may pose additional rules, such as restricting commercial use, but may
not do anything prohibited by this license regardless of how much you have changed.
Derivative software must also include credits for Kaishikuon, the developer of
NoLyui’s DiffSinger, as well as the supplemental data provided by CometP and the つ
くよみちゃん UTAU Database Unofficial Research Corpus in addition to NoLyui and
IsuelEasel.
Credit & Redistribution
Regardless of circumstance, redistribution requires prior permission from IseulEasel.
Credit is appreciated for all usage. Even when credit attribution is required, only the
name NoLyui is mandatory for general usage. Credits to the voicer, Iseul / IseulEasel,
the developer Kaishikuon, or the supplementary data contributors Kaishikuon,
CometP, and the つくよみちゃん UTAU Database Unofficial Research Corpus, are
appreciated but not required unless you are using the voice library for private research
or text-to-speech.
For the purposes of discoverability, if you have translated NoLyui’s name into another
language, please include the original name in roman characters as well


Iseuleasel