Synthesizer V

From Moegirlpedia
Jump to: navigation, search
爱莲娜芙缇info.jpeg
Moegirlpedia welcomes your assistance in improving this Synthesizer V-related article.

As you read this article, you're welcome to participate in editing this page.
Before editing, please read the wiki quickstart, edit guidelines and retrieve relevant information.
We wish you a good time on Moegirlpedia.

Synthesizer V R2 logo.png
Information
Product Type Voice synthesis software
Developer Kanru Hua
First Update Preview version
(August 19, 2018)
Latest Update Synthesizer V Studio 1.5.0
(December 23, 2021)
Languages English, Japanese, Chinese, Korean
Publisher Dreamtonics Co., Ltd.
Color 15E879


Synthesizer V (SynthV for short) is a singing synthesis engine developed by Dreamtonics Co,. Ltd. headed by Kanru Hua (V stands for the fifth architecture iteration). The engine uses the self-developed LLSM (low-level speech model) technology based on artificial neural networks and splicing synthesis algorithms, and can generate natural sounds using only a small amount of sampled data.

Synthesizer V uses a library with two different versions, Standard and AI, corresponding to WaveNet vocoder-enhanced traditional spliced vocal synthesis (Standard) and machine learning-based vocal synthesis (AI).[1]

Its engine and sound library software sales agent in Mainland China is now Quadimension (Beijing Photek Technology Development Co., Ltd.).

Development History

Editor

The first generation
  • On December 1, 2017, developer Kanru Hua (ID: sleepwalking) revealed SynthesizerV for the first time at Bilibili. Synthesizer V was the result of 7 years of research and development[2].Kanru Hua uploaded a follow-up video on May 7, 2018.[3].
  • Producer Kyo contributed "Although the singing is invisible" at Bilibili. On June 16th, Cillia submitted a cover of "silence" with the voice library ENG-F1 sound library Bilibili[4]. It was widely acclaimed for its very natural sound in the sound library and the divine adjustment of the two adjusters.
  • On August 19, 2018, the preview version of Synthesizer V was released, with the built-in ENG-F1 sound library-Eleanor Forte.
  • On December 7, 2018, Kyo submitted a cover of the song "Miss You" using Synthesizer V’s first Chinese voice library AiKO in bilibili[5]; in December 23, he released an authorized audition DEMO of the Japanese voice library GENBU.[6].
  • On December 24, 2018, developer Kanru Hua released the first official version of Synthesizer V [7], sold by Animen and Quadimension.
Second generation
  • On December 31, 2019, developer Kanru Hua released a preview version of Web Synthesizer V based on Synthesizer V R2[8], built-in Web version Japanese sound library Saki.
  • On April 11th, 2020, Stardust Official Weibo released an early demo track based on the Synthesizer V R2 sound library Chiyu remix of "404 Not Found". [9].
  • On June 26, 2020, Dreamtonics officially released Synthesizer V Studio[10], bringing a brand-new user interface and Synthesizer V Engine 2.0, sold by AH-Software, Quadimension and Animen.
  • On October 30, 2020, Dreamtonics submitted a Synthesizer V AI singing synthesis test song based entirely on deep learning[11].
  • On December 25, 2020, Synthesizer V AI was officially released, merged into the existing Synthesizer V Studio, and provided as a free upgrade.
  • On February 19, 2021, Synthesizer V AI updated its second-generation technology to improve rendering speed and sound quality, improve synthesis stability and reduce the size of the AI sound library.
  • On June 18, 2021, Synthesizer V AI ushered in the third-generation upgrade, bringing obvious detail enhancement and richer singing expressiveness [12].
  • On October 14, 2021, Dreamtonics submitted a Chinese AI female voice test song based on the fourth-generation Synthesizer V AI, and started a call for names. It was named Qingsu on November 5th.
  • On November 18, 2021, Synthesizer V AI officially brings the fourth-generation update, adding a customizable, sound library-independent automatic pitch adjustment function [13].
  • On November 23, 2021, SynthesizerV AI brought a fifth-generation update, audio fidelity and speed have been improved, and the editor also supports cross-lingual song synthesis from the AI voice library (requires a vocal library upgrade from the vocal library manufacturer, no additional purchase or download required to synthesize vocals in other languages)[14]

Released voice libraries

Synthesizer V
  • On August 19, 2018, Synthesizer V English sound library Eleanor Forte was released.
  • On November 11, 2018, the Synthesizer V Japanese sound library Yamine Renri became available for download.
  • On December 24, 2018, the Synthesizer V Japanese sound library GENBU became available for download.
  • On December 28, 2018, Animen released the Synthesizer V Chinese Sound Library AiKO, Quadimension also sold it as an agent.
  • On April 19, 2019, Quadimension released the Synthesizer V Chinese Sound Library Chiyu audition and opened pre-sale.
  • On May 31, 2019, Quadimension released the Synthesizer V Chinese Sound Library Cangqiong audition and started pre-sale.
  • On June 28, 2019, Quadimension released the Synthesizer V Chinese Sound Library Shian audition song and opened pre-sale.
  • On August 22, 2019, Quadimension released the Synthesizer V Chinese Sound Library Haiyi audition and opened pre-sale.
Synthesizer V Studio
  • On June 26, 2020, Quadimension released Synthesizer V Studio Chinese Sound Library Chiyu, Cangqiong, Shian and Haiyi and pre-sale started; AH-Software released Japanese voice libraries Saki and Kotonoha Akane & Aoi.
  • On July 13, 2020, Animen released Synthesizer V Studio Chinese sound library AiKO, Japanese sound library GENBU and Saki; 7 On 29th, agent sales of Chinese voice database Chiyu, Cangqiong, Shian and Haiyi.
  • On August 7, 2020, Quadimension released the Synthesizer V Studio Chinese Sound Library Muxin audition and pre-sale started.
  • On August 12, 2020, Quadimension will release the Synthesizer V Studio Chinese Sound Library Minus, and the pre-sale will start on August 28.
  • On December 25, 2020, Dreamtonics released the Synthesizer V Studio Japanese sound library Saki AI, which was also distributed by AH-Software and Animen.
  • On January 21, 2021, AH-Software sold Synthesizer V Studio Japanese sound library GENBU as an agent.
  • On March 18, 2021, AH-Software released Synthesizer V Studio Japanese sound library Koharu Rikka & Koharu Rikka AI.
  • On June 18, 2021, AH-Software released the Synthesizer V Studio Japanese/English sound library Tsurumaki Maki & Tsurumaki Maki AI.
  • On October 13, 2021, Animen released the Synthesizer V Studio Japanese sound library Yamine Renri, and opened a free registration channel.
  • On October 17, 2021, AUDIOLOGIE released the Synthesizer V Studio English Sound Library ANRI and pre-sale started. Animen also sold it as an agent.
  • On November 20, 2021, Dreamtonics opened the pre-sale of Synthesizer V Studio Chinese Sound Library Qingsu.
  • On November 26, 2021, AH-Software released the Synthesizer V Studio Japanese voice libraries Tsuina-chan and Tsuina-chan AI.
  • On December 1, 2021, Dreamtonics released the Chinese voice library Qingsu, and opened pre-sale in Taobao.
  • On December 6, 2021, Animen released the Synthesizer V Studio English voice library Eleanor Forte AI.
  • On December 11, 2021, AUDIOLOGIE released the Synthesizer V Studio English voice library ANRI
  • On January 19, 2022, Eclipsed Sounds released the Synthesizer V Studio English voice library SOLARIA.
  • On January 25, 2022, Quadimension released a demo song for Chinese voice library Infinity, presale started on January 27, and it was officially released on February 20.
  • On January 27, 2022, AH-Software released the Japanese voice libraries Kyomachi Seika and Kyomachi Seika AI.
  • On February 4, 2022, Dreamtonics released the Synthesizer V Studio English voice library Kevin and the Japanese voice library Ryo, these were also sold by AH-Software as an agent.
  • On April 9, 2022, Dreamtonics released the Synthesizer V Studio Chinese voice library Mo Zhen.
  • On May 4, 2022, Dreamtonics released the Synthesizer V Studio Chinese voice library An Xiao.

Technology

Official introduction

Functions

References

  1. ↑ The two sound libraries are significantly different in size. the AI sound library requires less than 100MB of storage space as it only stores the models trained for machine learning.
  2. ↑ sleepwalking. -{zh;zh-hans;zh-hant|新的歌声合成软件}-. Bilibili. 2017-12-01. 
  3. ↑ sleepwalking. -{zh;zh-hans;zh-hant|自制歌声合成软件Synth V展示阿童木主题曲}-. Bilibili. 2018-05-07. 
  4. ↑ 【新歌声合成引擎SynthV】SILENCE【自调教cover / cillia】
  5. ↑ 【SynthV首款正式声库抢先试用】好想你【艾可（AiKO）】
  6. ↑ 【SynthV】全新男声声库DEMO【ゲンブ（玄武）】
  7. ↑ sleepwalking. -{zh;zh-hans;zh-hant|歌声合成软件 Synthesizer V 正式发布.}-. Bilibili. 2018-12-24. 
  8. ↑ Web Synthesizer V
  9. ↑ SynthV第二代歌声合成引擎先行试听「404 Not Found」
  10. ↑ Synthesizer V Studio conference full playback
  11. ↑ Synthesizer V AI歌声合成测试曲
  12. ↑ 第三代 Synthesizer V AI 歌声合成技术
  13. ↑ /cv14059942 Synthesizer V Studio 1.4.0 更新日志
  14. ↑ Synthesizer V AI 跨语种合成更新 - Synthesizer V Studio 1.5.0 更新日志
Retrieved from "https://en.moegirl.org.cn/index.php?title=Synthesizer_V&oldid=38915"