ZETA (婕妲; Jiédá) is a feminine vocal for the DeepVocal engine. Developed by LainHikaru, she was released under the VOICEMITH label as their fourth voice synthesizer character.
History[]
2019[]
On July 27 2019, voicebank developer LainHikaru posted a screenshot of a private demonstration video using a prototype LanCee (蘭希) DeepVocal voicebank, labeled as an "official secret project".[1] In a reply to a comment on the post, LainHikaru noted that the release at the time was unsure, however there was an official plan for the vocal to appear in a new program. In a second comment, LainHikaru noted that there was a group internal test as opposed to a public beta test of the vocal. Further details were not given.
2020[]
Throughout February 2020, VOICEMITH posted stories featuring "Ashley" and her search and development of future VTubers and vocalists, including codenames of "VB01" and "VB06".[2][3][4] In this story, there was mention of a Virtual Idol Project and six recorded voice libraries at the time, with hints that there was a switch in synthesizer engines. To introduce female singers, VB06 was recommended as a voice library though VB01 was a tentative choice.[5] On March 18, VB06's voice was officially revealed and confirmed to be able to sing in Mandarin Chinese using the DeepVocal voice synthesizer.[6] On March 24, it was decided to move ahead with VB06.[7] On March 25, her first Japanese demo was released, confirming her to be a bilingual vocalist.[8]
On April 14, another story was released, revealing that there was at least another month to perform internal tests of the voice library. Regarding the commercial distribution, it was mentioned that there were plans, but it was decided that if it was free, people would prefer to use it. If it was sold, people might not. After internal tests, external tests and demo songs were planned, as well as illustrations to use as reference points.[9] On April 21, another story revealed that the chosen illustrator would be busy and that it may take another three months before they were available. There were no intentions to switch to another illustrator and VOICEMITH chose to wait.[10] June 10's story revealed the brainstorming of the vocalist's name, which was said to be a difficult process when thinking of one out of nothing. There was no specification of whether the name should be in Chinese or English and the character illustration had not been received yet, revealing a fear that the character's image might not match the name. While everyone discussed potential names, many strange ones were thrown in, and some people laughed and said that they should visit the numerology website first to count the number of strokes in a name to find out which ones were inauspicious. At that point, "Ashley" looked at the codename "VB06", taking note of the number "6" and thought of a potential name.[11]
On June 11, a female character appeared on the official VOICEMITH Facebook page's header image along with the word "ZETA" near her.[12] It was revealed that ZETA was a new singer making her debut with her first original song. Her illustrator was confirmed to be TaaRO and her voice provider was confirmed to be Rose. She was developed by LainHikaru. An official webpage for her was also launched which contained downloads for her DeepVocal voicebanks, marking her official release.[13] A Q/A was also posted, detailing the decision behind the name. Similar to the story posted in June, it was noted that ZETA's internal number was VB06 prior to launch. One of the team members was fixated on relating the name to the number "six", however, in the beginning, it was prioritized that the vocalist should have a Chinese name and the team was heatedly discussing names made up of two to four characters. Later, some of the members thought that the Greek alphabet was widely used in mathematics, physics, chemistry, and astronomy. These letters also sounded like names. It was regarded as simple and easy to read, so the sixth Greek letter "Zeta" (ζ) was chosen. However, since not everyone can read this, it was decided to use the English pronunciation as the name.[14]
On June 14, a second Q/A was posted in regards to why DeepVocal was selected for ZETA. VOICEMITH revealed that in the past, they tried to contact the developers of different voice synthesizers, both commercial and free. However, when considering the real conditions, they realized that there were not many people willing to pay for this library and they hoped to have more creators willing to use the library for music, so a free voice synthesizer became the biggest consideration. It was noted that though DeepVocal might not be the best, it was relatively stable and it wasn't complicated to use in addition to being free, which was the most appealing feature to VOICEMITH.[15] It was later posted that TaaRO was selected as ZETA's illustrator, who was informed in advance that her attribute was "DARKNESS", giving her a purple theme. VOICEMITH also proposed a concept similar to "elves" and gave TaaRO information about her personality and to design her based on these qualities.
On June 15, a third Q/A was released asking about potential easter eggs. VOICEMITH revealed that ZETA, and LanCee, one of the VTubers under their label, were voiced by the same person, Rose. They revealed that VB06 samples were recorded since 2018 and initially, UTAU was planned to be used as the engine. However, at the time, there was a growing trend for VTubers in Japan, rousing VOICEMITH's interest to create their own. After initial meetings and discussions with Rose, VOICEMITH decided to suspend the release of VB06 and launch LanCee instead. It was not until later that they decided to use the DeepVocal engine that they started to calibrate voice libraries. After evaluation, it was initially planned to release after the Lunar New Year, but due to the COVID-19 pandemic, the release was delayed again until June. When it was mentioned that LanCee and ZETA sounded different from each other, VOICEMITH noted this to be an interesting phenomenon. They noticed that ZETA's vocals were articulated differently and accidentally retained a lot of obvious nasal sounds which were not present in LanCee's voice. At the same time, this kind of strong nasal singing may sound intimate and comforting. According to VOICEMITH, nasally voices were considered rare among the current virtual singers and they hoped that her nasal voice would make people like her singing more.[16]
2022[]
On May 31, 2022, after XIA YU YAO's AI voice library production crowdfund launched, a fan asked if there would be AI voice libraries for the other VOICEMITH voice synthesizer characters (including OSCAR and JOAN, two UTAUs released in August 2016). The company replied "Possibly but not this year".[17]
2023[]
Some time after the release of XIA YU YAO's Synthesizer V AI library, VOICEMITH's website was modified to no longer include the other voice synthesizer characters. There was no announcement regarding the change to the website nor was a reason provided.[18]
2024[]
On May 10, 2024, it was announced that all UTAU and DeepVocal voicebanks produced by VOICEMITH would cease distribution on May 24. Free redistribution from other parties would not be prohibited, but selling the voicebanks was forbidden. VOICEMITH would also not assume any liabilities for any losses that may occur to users during these processes. The UTAU voicebanks of XIA YU YAO, OSCAR, and JOAN, as well as the DeepVocal voice for ZETA, were officially replaced with their ACE Virtual Singer AI versions due to the latter's intention of being faithful recreations of the original voicebanks.[19]
Character[]
ZETA is described by VOICEMITH as a girl who is good at observation and is delicate at heart. Innocent in nature, she is often experiencing loneliness and cannot recall memories before the age of 6. Her melancholy appearance hides her strong heart and her nasal voice brings friendliness.
All VOICEMITH singers have a designated "element" as an attribute. ZETA has the attribute of "Darkness", which is described as being able to take those who were born of darkness and negativity and then remake them as pure.
Etymology[]
"ZETA" is the sixth letter of the Greek alphabet (ζ) and has a value in the Greek mathematics system of 7. According to the stories prior to her release and her first Q/A, ZETA's name was specifically chosen in relation to the number "6" in her codename "VB06", which started from a team member's fixation on the number. When discussed, it was noted that Greek letters sounded like names and were simple and easy to read, hence the choice of her final name.
Her name is written phonetically in Chinese as "婕妲" (Jiédá).
Appearance[]
Going hand-in-hand with her elemental theme, ZETA has a darker color scheme of mostly purple and black with some white/grey highlights. Her design was based on her personality and according to VOICEMITH, she was to have a concept similar to "elves".
Relations[]
- XIA YU YAO (夏语遙) - Fellow VOICEMITH Singer with the "Wind" attribute.
- OSCAR (奧斯卡) - Fellow VOICEMITH Singer with the "Fire" attribute.
- JOAN (蕎安) - Fellow VOICEMITH Singer with the "Ice" attribute.
- LanCee (蘭希) - VOICEMITH V-Tuber with the "Light" attribute and shared voice provider
- Senior Doo (大嘟) - Fellow male VOICEMITH vocalist from the future, themed after "water".[20]
- Junior Doo (小嘟) - Fellow male VOICEMITH vocalist themed after "water" and the ancestor of Senior Doo.
Voicebank[]
Demonstrations |
---|
| |
---|---|
但願人長久 (Dàn Yuàn Rén Chángjiǔ) | YouTube |
aimai | YouTube |
邪惡的漆黑 (Xié'è de Qīhēi) | YouTube |
Download[]
Gallery[]
References[]
- ↑ https://t.bilibili.com/280839300079023411?spm_id_from=333.999.list.card_time.click
- ↑ https://www.facebook.com/voicemith/posts/pfbid0KjWLyRVTyyJZ5B6FAgo3SHKkVbduZx81vDT1xM6gLkzRLeSEdoL71MbgZddoNJJ4l
- ↑ https://www.facebook.com/notes/666433587612023/
- ↑ https://www.facebook.com/voicemith/posts/pfbid02hM8S9rqM9M6szmTMhfQRgDAYk3yGnH3zY8wwbcbMdAph8Ui7QzqJiVnowCUYNNkGl
- ↑ https://www.facebook.com/voicemith/posts/pfbid0n8Qy8RNBsRuLRQUkdMThJtyApu58BoufwxYZqs5psfbWQf7VvRKfagiuemNe81egl
- ↑ https://www.facebook.com/voicemith/posts/pfbid02V6PXRfNyUfE9Hx7zgjCwnhj8873jTyDasHrXBCfQ1mv9U3xMFhCTQAg3TA2jpLj7l
- ↑ https://www.facebook.com/voicemith/posts/pfbid02qrDG5fpgtxqm88SmsxWni86Rim5Ca9bPo28pvbPdzPr4S2GPfa9aA3cJipgvMqTHl
- ↑ https://www.facebook.com/voicemith/posts/pfbid032wgHZninrFJqfxxJhu2xehVsXoA2yyKE4uAEsH4JqHkPXPVs9QThqiXkWQscviHEl
- ↑ https://www.facebook.com/voicemith/posts/pfbid02TrCepw93fXapia5GXdoUQ8MeqBdFZZbsjECr4DtYbXdo8aVq3LPmyq3gDFCqQExml
- ↑ https://www.facebook.com/voicemith/posts/pfbid0zxyihQ2bdCQddCV5vanA62SmQ934qexC58bvca1mj7Ktv2qoreQuMaCAbsc5nRql
- ↑ https://www.facebook.com/voicemith/posts/pfbid0fk5k3B4RmVsamY35C2teVETFBbJujpGXR9whmy3UweM89jpBv6V2D8XYxyKYQbp1l
- ↑ https://www.facebook.com/voicemith/photos/a.643071735791640/2842995839132541/
- ↑ https://www.facebook.com/voicemith/posts/pfbid0rw1UfRhG94AwT75VpxZ3yguDkGHU448mE4XxDXdPszbhzftW4BaEHsXQ3t5YijFql
- ↑ https://www.facebook.com/voicemith/posts/pfbid035vmwsnLUUXmZhCw93W8tbuWoHDgcTLYJB8QhfPTqR2mrqTpAyv6aiwGTcwz822B7l
- ↑ https://www.facebook.com/watch/?v=3031921880221730
- ↑ https://www.facebook.com/voicemith/posts/pfbid0gRHiTw1X41DQV7Qw1mqsQsn1Nkei9YqYDH1HXaNib6DcKFZdr3iC8EDEruE3iKHxl
- ↑ https://twitter.com/YuyaoOfficial/status/1637750748409888769
- ↑ https://media.discordapp.net/attachments/695380056943493280/1085810360520560732/image.png - Screenshot of VOICEMITH's website as of March 15, 2022
- ↑ https://twitter.com/YuyaoOfficial/status/1788858002705146242
- ↑ https://twitter.com/JuxerOfficial/status/1781653364402786341
DeepVocal Voices |
---|