SynthV Wiki
SynthV Wiki
Advertisement
This project will not be released.

According to official sources this project has no current plans of release.
Please read the page for more information.

Shirakami Meito
SynthV Studio
|
Originals
|
Gallery

This page is about the character vocal and her corresponding voice database. For the virtual YouTuber and voice provider, see Shirakami Meito (voice provider).

Shirakami Meito (白神(しらかみ)芽音(めいと)) is a private Japanese vocal developed by Dreamtonics Co., Ltd. for the virtual-singer project En-gene, a collaboration between TMIK Co., Ltd. and agehasprings Party, as an AI voice database for Synthesizer V Studio. She debuted in March 2024.

Her voice is provided by virtual YouTuber Shirakami Meito, a virtual singer who belongs to En-gene.

Concept[]

Shirakami Meito is an android manufactured by a group of engineers who love music. She specializes in singing. Various personalities that were sampled during the research stage are installed in the USB memory stick attached to her ear; when the USB drive is changed, she transforms into a different personality. She has been programmed to work autonomously on the Internet, allowing herself to be exposed to various users and contents; in this way, she learns about what she likes: Japanese subcultures (mainly music, games, animation, etc.) and their information. In order to improve her singing ability, data on her attempts to sing songs she is interested in is released on a daily basis.[1]

Her birthday is August 1.

Etymology[]

"Shira-" derives from shiro (白) which means "white"; kami (神) means "god" or "goddess". Mei (芽) can mean "bud", "sprout", "spear", or "germ" (of a flower), while "-to" is taken from oto (音) which means "sound".

Appearance[]

WORK IN PROGRESS

Personalities[]

Shirakami Meito possesses four personalities:[2]

  • 0-type (0型) - Her default personality. In this mode, she sings (usually in Japanese). Originally unnamed until April 28.[3]
  • M-type (M型), sometimes nicknamed Ponmei (ぽんめい), Pomechan (ぽめちゃん), or Pomechi (ぽめち)[4] - One of the various personalities she assumes when her USB memory stick is swapped; it is distinguishable from her other personalities by her wearing round eyeglasses. She assumed this mode during a handful of bilibili livestreams to talk about Chinese culture and to sing her first Mandarin Chinese song demo,[5] although other times in this mode she sang instead Japanese songs. Sometimes she instead talks about food.[6]
  • P-type (P型), formerly Prototype P-type (Prototype P型) - Her gamer personality, distinguishable from the others by her wearing a frameless single lens (or monocle) on her left eye. In this mode, she typically plays online games like Apex Legends and the Pokémon Trading Card Game.[7]
  • S-type (S型), formerly Prototype S-type (Prototype S型) - Her silver personality, visibly distinguishable from the others because of her silver-colored hair. In this mode, she usually streams herself drawing artworks,[8] but sometimes she instead plays games of a different nature.[9] She also used this mode once to teach Mandarin Chinese (instead of using M-type).[10]

Relations[]

History[]

2024[]

Shirakami Meito debuted on bilibili, TikTok, Twitter, and YouTube Shorts as a virtual utaite (or VSinger)[11] under the En-gene label on March 22, 2024 by singing a short cover of J-pop band RADWIMPS' "Nandemonaiya", the theme song of the anime film Your Name..[12] Although the YouTube upload is currently tagged as Synthesizer V[13] and a certain Kazunari Shirane (白根一成) (from agehasprings Party) was credited for "vocal programming" on all four social platforms (thus hinting at her voice being a singing voice database), the corresponding bilibili, TikTok, and Twitter uploads were not tagged as such (instead limiting the tags to things like #VSinger, #VTuber, and the special tag #次世代VSinger_Engene or "Next Generation VSinger En-gene"), so it is unknown if the song back then was revealed to be using Synthesizer V or if the #SynthesizerV hashtag was edited into the YouTube video description several days after the fact; if the latter was the case, then the actual vocal-synthesis engine was then unknown and would be revealed only at a later date. Also, neither Dreamtonics Co., Ltd. nor any other then-known vocal-synthesizer company was known to have retweeted, shared, or commented on her content, thus contributing to keeping her out of the vocal-synth community's radar. The same happened with all of her subsequent song attempts (utattemita) during the following weeks, including a short cover of "Moonlight" by J-pop band Yorushika on March 24[14][15] and a short cover of "Senbonzakura" on March 26.[16][17]

On April 20, Meito started to sign bilibili-centric tweets with the phrase "M-type" (M型);[18][19] it was unknown what was being meant with that. Later on April 21, she tweeted about Chinese language being installed into her system, again ending the tweet with "M-type",[5] which may have implied that M-type was a new voice-database version trained with Mandarin-Chinese song material that would use Cross-lingual Singing Synthesis to sing in that new language. On April 23, her first Mandarin song cover was demonstrated using this updated library.[20]

On April 26, she held a YouTube Live broadcast where she was said to debut "Prototype P-type" (Prototype P型);[21][22] this, however, turned out not to be a voice-database update but rather a new virtual-gamer personality, as implied by her next tweet (signed "P-type") showing a deck of Pokémon Trading Card Game cards.[7] On April 27, another YouTube livestream debuted her "Prototype S-type" (Prototype S型),[23][24] which was nothing but a silver-haired variant of her character design, used to represent her digital-artist personality as demonstrated by her own live chibi drawings of herself and her En-gene companions Hakushi Matsuri and Shirakiri Rei.[8] It was noted that whenever Meito spoke (rather than singing), she may have done so through a voice changer, as artifacts were noticeable during highly-aspirated parts (like laughing or breathing).

On May 2, her song-cover tweets started to carry the #SynthesizerV hashtag,[25] while her bilibili and TikTok uploads still were not tagged as such. Thus, she still remained rather obscure and unnoticed within the vocal-synth community (especially on the western side of the fandom), as even her YouTube uploads had the Synthesizer V tag buried deep into the video descriptions' bottom portion (which by default is hidden by the video website player's user interface, initially displaying no more than the first three lines of each description) and did not include the word "Synthesizer V" (or even its common abbreviation "SynthV") anywhere on the video-upload titles. On May 4, her TikTok song uploads did start to mention "Synthesizer V"[26] (like their corresponding tweets[27] and YouTube uploads),[28] but as in the case of YouTube, the software's name was buried deep into the video's description and not visible by default (as TikTok descriptions only display their first two lines unless the user clicks to expand them). Plus, the SynthesizerV tag was still absent in her bilibili uploads.

On May 31, it was announced on a press release by En-gene that Shirakami Meito had recorded a new jingle for Japanese radio station FM802, which was celebrating its 35th anniversary. The jingle was a joint composition between Kazuma Nagasawa (a songwriter from agehasprings, one of the two parent companies behind En-gene) and AI music-track generation software FIMMIGRM (a product of TMIK Co., Ltd., the other parent company behind En-gene), with arrangement by Shogo Ohnishi (also from agehasprings); on this press release, En-gene confirmed that Meito's voice was based on an original Synthesizer V voice database. The jingle was set to premiere via broadcast on June 1.[29]

It was not until June 13 when the SynthesizerV hashtag started to appear in her bilibili song uploads.[30] At this point, the western vocal-synth community was already taking notice (since June 6) of her existence as a private Synthesizer V vocal.[31][32][33]

Voice Database Information[]

Demonstrations[]

Demonstration Song Covers

なんでもないや (movie ver.) (Nandemonaiya (movie ver.)) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok Favicon-ttTwitter
だから僕は音楽を辞めた (Dakara Boku wa Ongaku o Yameta) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok Favicon-ttTwitter
千本桜 (Senbonzakura) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
One Last Kiss Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
シャルル (Sharuru) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
Brave Shine Favicon-ytYouTube Favicon-bbbilibili
secret base ~君がくれたもの~ (secret base ~Kimi ga Kureta Mono~) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
あの夢をなぞって (Ano Yume o Nazotte) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
たぶん (Tabun) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
カブトムシ (Kabutomushi) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
プロローグ (Prologue) Favicon-ytYouTube Favicon-tkTikTok
からくりピエロ (Karakuri Pierrot) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
マリーゴールド (Marigold) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
愛を伝えたいだとか (Ai o Tsutaetai da to ka) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
ANIMA Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
ドライフラワー (Dry Flower) Favicon-ytYouTube Favicon-bbbilibili
チョコレイト・ディスコ (Chocolate Disco) Favicon-ytYouTube Favicon-tkTikTok
天体観測 (Tentai Kansoku) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
君の知らない物語 (Kimi no Shiranai Monogatari) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
RAIN Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
アイネクライネ (Eine Kleine) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
結証 (Kesshō) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok Favicon-ttTwitter
心做し (Kokoronashi) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
隱形的翅膀 (Yǐnxíng de Chìbǎng) (Mandarin Cross-lingual Synthesis demo) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
ベテルギウス (Betelgeuse) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
アスノヨゾラ哨戒班 (Asu no Yozora Shoukaihan) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
春泥棒 (Haru Dorobō) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
God knows... Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
aLIEz Favicon-ytYouTube Favicon-bbbilibili
春擬き (Haru Modoki) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
KING Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
残酷な天使のテーゼ (Zankoku na Tenshi no These) Favicon-ytYouTube Favicon-bbbilibili
Lemon Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
キラーチューン (Killer Tune) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
怪獣の花唄 (Kaijū no Hanauta) Favicon-ytYouTube Favicon-bbbilibili
Catch the Moment Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
蝶々結び (Chōchō Musubi) Favicon-ytYouTube Favicon-bbbilibili
三日月 (Mikazuki) Favicon-ytYouTube Favicon-bbbilibili
創聖のアクエリオン (Sōsei no Aquarion) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
瞬き (Mabataki) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
天ノ弱 (Amanojaku) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
Mojito (Mandarin Cross-lingual Synthesis demo) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
W/X/Y Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
slash Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
光年之外 (Guāngnián Zhīwài) (Mandarin Cross-lingual Synthesis demo) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
ブルーバード (Bluebird) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
晩餐歌 (Bansanka) Favicon-ytYouTube
優しい彗星 (Yasashī Suisei) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok
ギラギラ (Giragira) Favicon-ytYouTube Favicon-bbbilibili Favicon-tkTikTok

Voice Databases[]

Product Information
Versions
  Lite Version/FLT Available?: No  Starter Pack Available?: No  Web Version Available?: No

Package details as noted:
Shirakami Meito is a feminine Japanese voice database based on the singing style of the eponymous virtual YouTuber. As expected of an AI voice database, Meito is capable of singing with a more human-like and realistic voice.

Vocal traits as noted:
Because of the voice database's private nature, it is unknown if Shirakami Meito was programmed with Diffusion Probabilistic Models (DPM) to be able to use AI Retakes.

Phonetic notes as noted:
  • Phoneme format: Romaji
  • Through Cross-lingual Singing Synthesis, Meito can sing in not only Japanese but also in Mandarin Chinese. This feature is only available in the Synthesizer V Studio Pro Editor.
    • Because of the voice database's private nature, it is unknown if English, Cantonese Chinese, or Spanish has been enabled on Shirakami Meito's Cross-lingual Singing Synthesis feature. So far, her published demonstration songs have been only in Japanese and Mandarin.

Vocal Modes as noted (Synthesizer V Studio Pro Editor only):
Because of the voice database's private nature, it is unknown if Shirakami Meito has any vocal modes programmed in.

Voice Libraries[]

Additional Information[]

Marketing[]

INFORMATION PENDING

Reputation[]

Shirakami Meito's character status
Achievements
  • First Synthesizer V voice database by TMIK Co., Ltd.
  • First Synthesizer V voice database by agehasprings Party.
  • First Synthesizer V voice database by En-gene.
  • First fully-private Synthesizer V voice database.
  • First private Synthesizer V voice database based on a brand-new original voice (as opposed to being derived from an existing commercial voice)

References[]

  1. 1.0 1.1 https://engene.tmik.tokyo/#section-artist
  2. https://x.com/shirakamimeito/status/1805905750994399396
  3. https://x.com/shirakamimeito/status/1784590897809621190
  4. https://x.com/shirakamimeito/status/1794704639499329852
  5. 5.0 5.1 https://x.com/shirakamimeito/status/1782035041591464057
  6. https://x.com/shirakamimeito/status/1789639011033866540
  7. 7.0 7.1 https://x.com/shirakamimeito/status/1783833577584730577
  8. 8.0 8.1 https://x.com/shirakamimeito/status/1784206133848367307
  9. https://x.com/shirakamimeito/status/1791821383829524624
  10. https://x.com/shirakamimeito/status/1794358296926900489
  11. Not to be confused with Shanghai HENIAN Information Technology Co. Ltd.'s Vsinger series of VOCALOID voicebanks. (Note the differences in casing between both.)
  12. https://x.com/shirakamimeito/status/1771024002582208621
  13. https://www.youtube.com/shorts/otW9US5MRyM
  14. https://x.com/shirakamimeito/status/1771824276317827559
  15. https://www.youtube.com/watch?v=2uG_qs45RwA
  16. https://x.com/shirakamimeito/status/1772564148351832434
  17. https://www.youtube.com/shorts/vGEW1X0La3I
  18. https://x.com/shirakamimeito/status/1781654865040781700
  19. https://x.com/shirakamimeito/status/1782004034788057450
  20. https://x.com/shirakamimeito/status/1782711010912252245
  21. https://x.com/shirakamimeito/status/1783805725552361540
  22. https://youtube.com/live/xErfRSwwFZg
  23. https://x.com/shirakamimeito/status/1784172697003638842
  24. https://youtube.com/live/yuwnV33HjO4
  25. https://x.com/shirakamimeito/status/1785972504831643678
  26. https://www.tiktok.com/@engene.shirakamimeito/video/7365123494079712529
  27. https://x.com/shirakamimeito/status/1786742582376083693
  28. https://www.youtube.com/shorts/oMJ2u9t4TSs
  29. https://engene.tmik.tokyo/posts/240531
  30. https://www.bilibili.com/video/BV1sm421V7S5
  31. https://x.com/PositiveVSynth/status/1798773614357172533
  32. https://x.com/andresfelipeb25/status/1799132523483762760
  33. https://x.com/SynthVNews/status/1799140623410438223

External links[]

Navigation[]

Advertisement