SynthV Wiki
Advertisement
Koharu Rikka
SynthV Studio
VOICEROID
CeVIO AI
VoiSona Talk
VOICEPEAK
|
Originals
|
Gallery

This article is about the Synthesizer V Studio software known as a voice database. If you are looking for the Synthesizer V character then click here.

History[]

When TOKYO6 ENTERTAINMENT was first established in December 2019, at the time, the company's president, Ryuichi Akasako, could not suddenly make a living solely through this company, so he began work as a programmer at the same time and made use of his savings. He had some idea of what he wanted to do when he started the company, thus deciding to launch a crowdfund for TOKYO6 ENTERTAINMENT's first vocalist, Koharu Rikka, while consulting with AH-Software Co. Ltd..[1]

2020[]

A crowdfunding campaign was launched on May 28, 2020 to fund for Koharu Rikka's VOICEROID2 voicebank and its stretch goals, which included the final stretch goal for 25 million yen: a singing voicebank.[2] This amount was reached on June 6, about a week and a half after the campaign began, but the actual synthesizer was not specified at the time.[3] On June 26, Synthesizer V Studio was revealed as her singing engine.[4] On August 8, it was announced that recording for her Synthesizer V bank started.[5]

On October 30, Rikka was announced that she would be developed for the CeVIO AI voice synthesis engine for speech. For singing, she was confirmed to have not only a Standard voice database but also the newly introduced AI type of voice database. She was slated for a Spring 2021 release.[6] [7]. In a press release from TOKYO6 ENTERTAINMENT, it was announced that the reason why Rikka was in production for CeVIO AI was due to the fact that it also used AI technology, unlike VOICEROID.[8]

2021[]

On February 18, 2021, during an AH-Software Co. Ltd. Livestream, Rikka's Synthesizer V AI voice database was showcased.[9][10] Following the stream, AH-Software announced a special campaign on Twitter allowing users to apply via DM to receive a physical copy of Synthesizer V Studio Pro and Synthesizer V Koharu Rikka AI Complete.[11]

Rikka officially released on March 18 for both CeVIO AI and Synthesizer V Studio. Lite versions of the Synthesizer V voice databases were also released on the same day via AH-Software's website.[12][13]

2022[]

On March 24, 2022, Rikka was confirmed to be receiving a physical box for only her AI voice database on April 13. This was due to high demand for AI-type voices.[14] On March 26, the release of Synthesizer V Studio version 1.6.0 brought new features: Vocal Mode and Instant Mode. TOKYO6 announced that these features would be brought to Rikka and Natsuki Karin later due to the need of making adjustments to the voice databases.[15] On May 20, it was announced on TOKYO6's livestream that Rikka would be receiving a larger-scale update to match Karin's and Hanakuma Chifuyu's quality. This would include additional recordings to improve her range with Vocal Mode.[16] The updates covered the following:

  • Improved voice quality by updating the acoustic model
  • More faithful to the original recordings' pitch
  • Voice was clearer and connections were smoother
  • Pronunciation and vowels were made clearer
  • Support English, Mandarin Chinese, and Japanese
  • Improvements made to the Auto Pitch Function

It was also noted that additional songs recorded by Yoshino Aoyama were added to Rikka's data which was necessary for Vocal Mode.[17] In addition, the vibrato of Rikka's original AI release was reported to be weak, so it was improved in the new recordings.[18]

On July 21, Rikka was one of the only AI voice databases not to receive an update for HDVM compatibility (along with Xingchen Infinity).[19] Similarly, she was also the only AH-Software Co. Ltd. AI voice database to not receive Vocal Modes in this update. However, her official Twitter account assured fans that they were still in development and would be implemented at a later date.[20]

On October 14, when asked about the status of Rikka's vocal modes, Akasako said that it would be released along with other improvements later.[21] On November 14, after the release of Rikka's v122b1 installer, TOKYO6 ENTERTAINMENT released a survey for its feedback.[22] On November 22, the company noted that the beta version would be used as the official update after receiving mostly positive feedback.[23]

Updates[]

AI - Full Version[]

Koharu Rikka AI v102 - March 24, 2021[24] 
Better pronunciation of rare phonemes and minor performance enhancements.
Koharu Rikka AI v105 - June 18, 2021[25][26] 
This free update supports the 3rd generation of AI singing voice synthesis.
Koharu Rikka AI v113 - December 16, 2021[27] 
This free update improves audio characteristics and synthesis quality.
Koharu Rikka AI v115 - December 23, 2021[28] 
A free update for Koharu Rikka AI was released following the Synthesizer V Studio 1.5.0 update. This enables her to support Cross-lingual Singing Synthesis.
Koharu Rikka AI v122b1 - November 10, 2022[29][30] 
A free update for Koharu Rikka AI was released following the Synthesizer V Studio 1.8.0b1 update. This enables her to use the new Diffusion Probabilistic Models (DPM) feature, as well as giving her access to the Vocal Modes feature previously introduced by the SVS 1.6.0 update. Her Vocal Modes are named: "Kawaii", "Soft", "Pops", "Emotional", and "Ballade". This update was confirmed to be using additional singing data recorded by Yoshino Aoyama, allowing Rikka AI to sing more realistically. It also improved the implementations of vibrato, which previous versions of Rikka AI had difficulty doing.
Koharu Rikka AI v122 - November 24, 2022[31][32] 
Final version update compatible for Synthesizer V Studio 1.8.0 which enables Koharu Rikka AI to use the new Diffusion Probabilistic Models (DPM) and Vocal Mode features if the beta version was not installed previously.
Koharu Rikka AI v123b1 - June 21, 2023[33][34] 
A free update for Koharu Rikka AI was released following the Synthesizer V Studio 1.9.0 update. This implemented the new rap feature available for the English and Mandarin languages. Cantonese Chinese was added to the Cross-lingual Singing Synthesis feature as well as general improvements to the overall function, resulting in more accurate pronunciation that sounds closer to a native speaker when singing foreign words.
Koharu Rikka AI v123 - July 6, 2023[35][36] 
Final version update compatible for Synthesizer V Studio 1.9.0 which enables Koharu Rikka AI to use the rap feature (English and Mandarin Chinese only) and allows for Cantonese Chinese Cross-lingual Singing Synthesis if the beta versions were not installed previously.
Koharu Rikka AI v124b1 - August 2, 2023[37][38][39][40] 
A free update for Koharu Rikka AI was released following the Synthesizer V Studio 1.10.0b1 update. This enhanced pitch generation by utilizing Reinforcement Learning from Human Feedback (RLHF), translating to fewer off-key errors, making better use of vibrato, and creating more expressive transitions between notes. This also allowed for the use of the new "Enhancement" slider in the AI Retakes panel, which controlled the level of changes introduced by RLHF. For Synthesizer V Studio Pro users, feedback submission was enabled, allowing users to consensually share their preferred takes with only Dreamtonics for research and development purposes, where models trained with this feedback would be deployed in future voice database updates. Users may also opt out of this feature if undesired.
Koharu Rikka AI v124b2 - September 28, 2023[41][42][43] 
A free update for Koharu Rikka AI was released following the Synthesizer V Studio 1.10.0b2 update, which was the second beta version following version 1.10.0b1 that introduced enhancements to voice databases and integrated the latest user feedback via Reinforcement Learning from Human Feedback (RLHF). Regarding the editor, numerous reported issues were addressed for a more fluid experience. A new generation of pitch models were trained using the extensive feedback from users over the past month, and the new models offered greater accuracy in voice timbre reproduction, as well as a reduction in mispronunciations. Dreamtonics also focused on issues from the previous beta release that made Vocal Modes seem less distinct on certain vocals, ensuring clearer differentiation in the 1.10.0b2 update. The latest update also introduced a new feature in the Note Properties Panel that allowed for even division of a note into multiple syllables, proving to be a useful tool for songs with faster tempos.
Koharu Rikka AI v126b1 - November 24, 2023[44][45][46] 
A free update for Koharu Rikka AI was released following the Synthesizer V Studio 1.11.0b1 update. This implemented support for the Japanese language in the rap feature. In addition, it disables Reinforcement Learning from Human Feedback (RLHF) enhancement for pitch generation by default, as requested by TOKYO6 ENTERTAINMENT.[47][48]
Koharu Rikka AI v126 - April 4, 2024[49][50][51] 
Final version update compatible for Synthesizer V Studio 1.11.0 which disables RLHF by default and enables Koharu Rikka AI to rap in Japanese, if the beta version was not installed previously. In addition, this update implements support for singing in the Spanish language in the Cross-lingual Singing Synthesis feature.

AI Lite[]

Koharu Rikka AI Lite v101 - June 18, 2021[52][53] 
A free update for Koharu Rikka AI Lite was released to coincide with the Synthesizer V Studio 1.3.0 update.

Voice Database Information[]

Demonstrations[]

Synthesizer V Studio Standard[]

Demonstrations

Demo Crossfade (AIのカタチ (AI no Katachi), The Everlasting Guilty Crown, オドッテ☆モロテ (Odotte ☆ Morote)) Favicon-nnNicoNico Favicon-ytYouTube
AIのカタチ (AI no Katachi) Favicon-nnNicoNico Favicon-ytYouTube Favicon-bbbilibili
オドッテ☆モロテ (Odotte ☆ Morote) Favicon-nnNicoNico Favicon-ytYouTube Favicon-bbbilibili
MIRA Favicon-nnNicoNico Favicon-ytYouTube Favicon-bbbilibili
The Everlasting Guilty Crown Favicon-nnNicoNico Favicon-ytYouTube
シャドーマンショー (Shadowman Show) Favicon-nnNicoNico Favicon-ytYouTube Favicon-bbbilibili
ローファー (Loafer) Favicon-nnNicoNico Favicon-ytYouTube

Synthesizer V AI[]

Demonstrations

あの星を探して (Ano Hoshi o Sagashite) Favicon-nnNicoNico Favicon-ytYouTube
あの星を探して (Ano Hoshi o Sagashite) - Short version w/ Automatic Pitch function Favicon-nnNicoNico Favicon-ytYouTube
ミゾオチ (Mizoochi) Favicon-nnNicoNico Favicon-ytYouTube Favicon-bbbilibili
AIのカタチ (AI no Katachi) Favicon-nnNicoNico Favicon-ytYouTube Favicon-bbbilibili
大腿四頭筋 (Daitai Shitoukin) Favicon-nnNicoNico Favicon-ytYouTube Favicon-bbbilibili
ジコショーカイ (Jikoshoukai) Favicon-nnNicoNico Favicon-ytYouTube Favicon-bbbilibili
已己巳己 (Ikomiki) Favicon-nnNicoNico Favicon-ytYouTube Favicon-bbbilibili
デラシネ (Déraciné) Favicon-nnNicoNico Favicon-ytYouTube Favicon-bbbilibili
デジタル$ゴールド (Digital $ Gold) Favicon-nnNicoNico Favicon-ytYouTube Favicon-bbbilibili
マジョニナル (Majo ni Naru) Favicon-nnNicoNico Favicon-ytYouTube Favicon-bbbilibili
追憶ポエジー (Tsuioku Poésie) Favicon-nnNicoNico Favicon-ytYouTube

Voice Databases[]

Product Information
Optimum
  Range: A3 ~ G5  Pitch Group: C4, E4, G#4, B4[54]

Versions
  Lite Version/FLT Available?: Yes (Lite)  Starter Pack Available?: Yes  Web Version Available?: No

Package details as noted:
Rikka is a feminine Japanese voice database that is designed to be suitable for all genres. She has a bright and energetic voice that has solid vocal support, and can also be breathy and nuanced.[55]

Vocal traits as noted:
  • Rikka's voice was advertised to be versatile.
  • Rikka was noted to have a rich middle range.[56]

Phonetic notes as noted:
  • Phoneme format: Romaji
  • ExVOICE voice files are included with Rikka's CeVIO AI Talk Voice or Synthesizer V AI voice database. There are over 1455 audio files, including binaural recordings, to use for a wide variety of genres.

Software issues as noted:
As the Lite voice databases are generally restricted to monopitch (one pitch) only, Rikka's range is noted to be E4 in this version.

System requirements as noted:

Hardware Requirements:

  • CPU: Intel Core i3 or higher
  • RAM: 2GB or higher
  • Disk Space: 1 GB or higher (for installing one voice database)
    • The amount of space required depends on the number of voice databases installed.
  • Display: 1280 x 800 or higher

Software Requirements:

  • Operating System:
    • Windows 10/8.1
    • Mac OS X: 10.11 or higher
    • Linux Ubuntu 16.4 or higher

Other:

  • DVD-ROM Drive (physical edition only)
  • Sound card
  • Internet connection is necessary for activation and obtaining updaters.

License agreements as noted:
  • TOKYO6 ENTERTAINMENT's Character Usage Guidelines - Japanese
  • AH-Software's End User License Agreement - Japanese
  • AH-Software's End User License Agreement - English
  • Regarding the Lite version:
    • AH-Software does not accept technical inquiries regarding free software nor do they guarantee or support the operation of all free software.
    • AH-Software is not liable for any damages (including loss of data) caused by using free software. Users are cautioned to use the free software at their own risk.
    • The program may not be transferred, sold, lent, distributed or rented without AH-Software's permission, regardless of whether it is paid or free of charge.
    • Reverse engineering, including modification, decompilation, and disassembly, of their program for any purpose is prohibited.
    • Koharu Rikka Standard Lite may not be used commercially in any way
    • When using the Lite version of this voice database, "This work uses the Lite version" must be mentioned on any uploads.

Voice database sample

Product Information
Optimum
  Range: G3 ~ F#5

Versions
  Lite Version/FLT Available?: Yes (Lite)  Starter Pack Available?: Yes  Web Version Available?: No

Package details as noted:
Rikka is a feminine Japanese voice database that is designed to be suitable for all genres and is based on the singing style of Yoshino Aoyama. She has a bright and energetic voice that has solid vocal support, and can also be breathy and nuanced. As expected of an AI voice database, this enables Rikka to sing with a more human-like and realistic voice.

Vocal traits as noted:
  • Rikka's voice was advertised to be versatile.
  • Rikka was noted to have a rich middle range.
  • As per the Synthesizer V Studio 1.7.0 update, through High Dynamics Voice Models (HDVM), which later migrated to Diffusion Probabilistic Models (DPM) per the 1.8.0b1 update on November 10,[58] AI voice databases may use AI Retakes to create multiple "takes" of a sung section with variations in pitch, dynamics, and timbre. This replicates the randomness of human vocalists, adds a more realistic feel, and allows users to find an ideal vocal without the need for "laborious tuning." Rikka did not have access to HDVM at the time but was later updated to use DPM since November 10 via her v122b1 update. This feature is only available in the Synthesizer V Studio Pro Editor.[59][60]
    • This does not apply to the Lite version.

Phonetic notes as noted:
  • Phoneme format: Romaji
  • ExVOICE voice files are included with Rikka's CeVIO AI Talk Voice or Synthesizer V AI voice database. There are over 1455 audio files, including binaural recordings, to use for a wide variety of genres.
  • On December 23, 2021, with the announcement of Synthesizer V Studio 1.5.0, all AI voice databases would be able to use Cross-lingual Singing Synthesis. This allows Rikka AI to be able to sing in not only Japanese but also in English and Mandarin Chinese using this new feature.[62] Since the release of her v123b1 update on June 21, 2023, she can also sing in Cantonese Chinese with this feature; since her v126 update on April 4, 2024, she can also sing in Spanish with this feature. Cross-lingual Singing Synthesis is only available in the Synthesizer V Studio Pro Editor.
    • This does not apply to the Lite version, thus restricting the voice database to Japanese only.

Vocal Modes as noted (Synthesizer V Studio Pro Editor only):
  • As of November 10, 2022, the paid version of Koharu Rikka AI became compatible with the Vocal Mode function introduced in Synthesizer V Studio 1.6.0. This was made possible through a larger-scale update to the voice database in order to match Natsuki Karin's and Hanakuma Chifuyu's quality and to improve Rikka's Vocal Mode range.[63] Her v122 update included five variations:
    • "Kawaii": Has a cute voice quality.
    • "Soft": Has a soft voice quality.
    • "Pops": Has a strong presence suitable for the bright and strong beats of Pop genres of music.
    • "Emotional": Ideal for singing emotionally and noted to make the best use of Yoshino Aoyama's vocal characteristics.
    • "Ballade": Suitable for Ballad genres.

System requirements as noted:

Hardware Requirements:

  • CPU:
    • Intel Core series 4th generation i5 (i5-4xxx) or higher
    • AMD Athlon X4 845, Ryzen series or higher
  • RAM: 2GB or higher
  • Disk Space: 1 GB or higher (for installing one voice database)
    • The amount of space required depends on the number of voice databases installed.
  • Display: 1280 x 800 or higher

Software Requirements:

  • Operating System:
    • Windows 10/8.1
    • Mac OS X: 10.11 or higher
    • Linux Ubuntu 16.4 or higher

Other:

  • DVD-ROM Drive (physical edition only)
  • Sound card
  • Internet connection is necessary for activation and obtaining updaters.

License agreements as noted:
  • TOKYO6 ENTERTAINMENT's Character Usage Guidelines - Japanese
  • AH-Software's End User License Agreement - Japanese
  • AH-Software's End User License Agreement - English
  • Regarding the Lite version:
    • AH-Software does not accept technical inquiries regarding free software nor do they guarantee or support the operation of all free software.
    • AH-Software is not liable for any damages (including loss of data) caused by using free software. Users are cautioned to use the free software at their own risk.
    • The program may not be transferred, sold, lent, distributed or rented without AH-Software's permission, regardless of whether it is paid or free of charge.
    • Reverse engineering, including modification, decompilation, and disassembly, of their program for any purpose is prohibited.
    • Koharu Rikka AI Lite may not be used commercially in any way
    • When using the Lite version of this voice database, "This work uses the Lite version" must be mentioned on any uploads.

Voice database sample

References[]

  1. https://www.dtmstation.com/archives/60921.html
  2. https://twitter.com/tokyo6info/status/1266374943165431809
  3. https://twitter.com/tokyo6info/status/1269628898637053956
  4. https://twitter.com/tokyo6info/status/1276370101172760576
  5. https://twitter.com/s_akasakov/status/1292095217659568129
  6. https://twitter.com/ahsoft/status/1322055701938622464
  7. https://www.ah-soft.com/press/synth-v/20201030.html
  8. https://note.com/tokyo6/n/nfd646a7bf86d
  9. https://youtu.be/n5vIp4TlCY0
  10. https://twitter.com/ahsoft/status/1362573316913991681
  11. https://twitter.com/ahsoft/status/1362387362785845249
  12. https://www.ah-soft.com/trial/synth-v.html
  13. https://twitter.com/ahsoft/status/1372383251231215617
  14. https://www.ah-soft.com/press/cevio/20220324.html
  15. https://twitter.com/tokyo6info/status/1507669520500273153
  16. https://www.youtube.com/watch?v=vvpaI_qIbS4
  17. https://twitter.com/uchuuzentai/status/1527603426103894016 TOKYO6's announcement about Rikka's update as translated by xuu.
  18. https://www.dtmstation.com/archives/60921.html
  19. https://twitter.com/dreamtonics_en/status/1550050403613835264
  20. https://twitter.com/rikka_info/status/1550061847948054528
  21. https://twitter.com/s_akasakov/status/1580836859315335168
  22. https://twitter.com/tokyo6info/status/1591981747180163072
  23. https://twitter.com/tokyo6info/status/1594852826085494784
  24. https://twitter.com/dreamtonics_en/status/1374918965328105473
  25. https://twitter.com/dreamtonics_en/status/1405722114452201472
  26. https://twitter.com/ahsoft/status/1405724392936509440
  27. https://twitter.com/ahsoft/status/1471405609740546049
  28. https://dreamtonics.com/en/cross-lingual-support-for-synthesizer-v-ai/
  29. https://dreamtonics.com/en/synthesizer-v-studio-1-8-0b1-update/
  30. https://note.com/tokyo6/n/n76d2f28e214a
  31. https://dreamtonics.com/en/synthesizer-v-studio-1-8-0-final-update/
  32. https://twitter.com/ahsoft/status/1595674717200916481/
  33. https://dreamtonics.com/synthesizer-v-studio-1-9-0-final-update/
  34. https://twitter.com/dreamtonics_en/status/1671427830629154816
  35. https://www.bilibili.com/read/cv24826703/
  36. https://twitter.com/ahsoft/status/1676865674898206720
  37. https://twitter.com/dreamtonics_en/status/1686648078571655168
  38. https://t.bilibili.com/825177687051993121
  39. https://dreamtonics.com/synthesizer-v-studio-1-10-0b1-update-enhancing-pitch-generation-with-user-feedback/
  40. https://twitter.com/ahsoft/status/1686653272474738689
  41. https://twitter.com/dreamtonics_en/status/1707304150059639108
  42. https://dreamtonics.com/synthesizer-v-studio-1-10-0b2-update/
  43. https://twitter.com/ahsoft/status/1707304906535956719
  44. https://dreamtonics.com/synthesizer-v-studio-1-11-0b1-update/
  45. https://twitter.com/ahsoft/status/1727970395691856025
  46. https://twitter.com/tokyo6info/status/1727987685288206599
  47. https://note.com/tokyo6/n/n742c974716b9 - The main reasoning behind TOKYO6's request to disable RLHF by default (which prevented the release of finalized versions of Koharu Rikka AI v124, Natsuki Karin v110, and Hanakuma Chifuyu v106 and also resulted in the permanent withdrawal of their last respective beta versions) was that the then-current implementation of RLHF learned AI-retake data selected as favorable by various different users of Rikka AI, Karin, and Chifuyu and then reflected it across all installations of their voice databases, thus affecting their overall preference trend; TOKYO6 believed that, due to things like variations in preferred vibrato swing size from user to user, etc., some users would get undesirable results out of their vocals, and in internal beta test results, TOKYO6 noted that the vocals' vibrato tended to be exaggerated and deviating from their character image. TOKYO6 was also concerned about RLHF results automatically affecting the default singing style (with no Vocal Modes applied) to a large extent, resulting in the decision to postpone any further updates to their voice databases until they were given the possibility to disable RLHF by default in the next update.
  48. https://twitter.com/tokyo6info/status/1717458061949026415
  49. https://dreamtonics.com/synthesizer-v-studio-1-11-0-update/
  50. https://twitter.com/ahsoft/status/1775800933538320621/photo/4
  51. https://twitter.com/tokyo6info/status/1775804166952403019
  52. https://twitter.com/ahsoft/status/1405724076769910792
  53. https://twitter.com/ahsoft/status/1405725227305213955
  54. https://www.ah-soft.com/synth-v/rikka/
  55. https://dreamtonics.com/en/synthesizerv/
  56. https://ubgoe.com/projects/200
  57. https://note.com/tokyo6/n/n90c4c901d596
  58. https://dreamtonics.com/en/synthesizer-v-studio-1-8-0-final-update/
  59. https://twitter.com/dreamtonics_en/status/1550048543624564736
  60. https://twitter.com/dreamtonics_en/status/1550048543624564736
  61. https://note.com/tokyo6/n/n90c4c901d596
  62. https://twitter.com/dreamtonics_en/status/1473941453810577410
  63. https://twitter.com/rikka_info/status/1590614998791122944

Navigation[]

Advertisement