CeVIO Wiki
Advertisement
🛠 This subject is work in progress.
Please bear with us while improvements are being made and assume good faith until the edits are complete.
For information on how to help, see the guidelines. More subjects categorized here.
🛠

CeVIO AI (チェビオエーアイ, pronounced che-ˈvē-ˈo) is the next generation of the CeVIO Creative Studio vocal synthesizer product developed in collaboration with Techno-Speech, Inc. capable of both speech and singing synthesis.

Overview

CeVIO AI was announced via Twitter on June 11, 2020[2] and is a continuation of the AI, incorporation for CeVIO first announced December 2018 by Techno-Speech, Inc. and Nagoya Institute of Technology.[3] CeVIO AI utilizes DNN speech synthesis by deep learning and recurrent neural network (RNN) to produce a more natural singing and speaking voice. [4]

Like it's predecessor, CeVIO AI will have a Song Editor and a Talk Editor, sold separately. CeVIO AI is noted to be different from CeVIO Creative Studio and that voicebanks from the previous versions are not compatible with the new software, however, support is being considered for future updates.[5]

With the announcement of Flower and Yuzuki Yukari's migration to CeVIO AI, Techno-Speech, Inc. noted that rather than viewing the move as an "evolution" from existing vocals, this should be viewed as an "expansion of possibilities" for voice synthesizer users.[6]

Development History

The following includes the official releases and major events only. More details can be found at the official webpage(in Japanese only).

  • On December 14, 2018 samples using AI incorporation were uploaded to Techno-Speech, Inc.[7]
  • IA TALK, ONE TALK, flower, and Yuzuki Yukari announced to migrate to CeVIO AI on July 22, 2020[8]
  • Demo and announcement of KAFU uploaded on October 10, 2020 [9]
  • Announcement of Tohoku Kiritan uploaded on October 21, 2020 [10]
  • CeVIO AI was formally announced via Twitter on October 29, 2020
  • Koharu Rikka was announced for CeVIO AI and Talk Editor demos uploaded to AH-Software Co. Ltd.'s YouTube channel on October 30, 2020 [11][12]
  • Logos for CevIO AI's package unveiled on November 5, 2020[13]
  • Demos of CeVIO AI's Song Editor featuring Sato Sasara were uploaded on November 20, 2020 [14][15]
  • On January 27, 2021 CeVIO AI Song Editor / Talk Editor version 8.0.9.0. was released for download.
  • On February 1, 2021, CeVIO AI version 8.0.11.0 was released, fixing several errors in the previous version.
  • On February 9, 2021 CeVIO AI version 8.0.13.0 was released.
  • On March 8, 2021, CeVIO AI version 8.0.9.0 was officially suspended.
  • On March 11, 2021, CeVIO AI version version 8.0.18.0 was released. Support was added for "CeVIO AI IA TALK" and "CeVIO AI ONE TALK"[16]
  • The day after, version 8.0.19.0 was released. Version 8.0.20.0 will be released few hours later to add support for Koharu Rikka's voicebank. Both of these versions fix critical bugs with song playing.
  • On March 16, 2021, CeVIO AI version 8.0.21.0 was released.
  • On April 1, 2021, CeVIO AI version 8.0.23.0 was released.
  • On July 15, 2021, CeVIO AI version 8.1.8.0 was released.
  • On August 18, 2021, CeVIO AI version 8.1.12.0 was released.
  • On August 27, 2021, CeVIO AI version 8.1.14.0 was released. Included talk voice improvements for IA TALK and ONE TALK
  • On October 21, 2021, CeVIO AI version 8.1.21.0 was released.

Speech Abilities

Like it's predessecor, CeVIO AI has access to a a Talk Editor. Talk Voice libraries are available for CeVIO AI as both Starter Packs (Talk Voice & Talk Editor) as well as Talk Voice libraries sold separate from the editor.

Singing Abilities

Like it's predessecor, CeVIO AI has access to a a Song Editor. Song Voice libraries are available for CeVIO AI as both Starter Packs (Song Voice & Song Editor) as well as Song Voice libraries sold separate from the editor.

Releases

Upcoming


Marketing

A preferential price for owners of CeVIO Creative Studio is being considered for upgrading, however it was noted there will be no free upgrades due to the time it has taken for development. Users have the options of purchasing CeVIO AI as a set of Talk Voice and Talk Editor, a set of Song Voice and Song Editor, or the Song Voice, Talk Voice and respective Editors separately..[17]

Trivia

WIP

System Requirements

  • Operating environment
    • Supported OS: Windows 10 Japanese edition (32 / 64bit)
    • Windows 8.1 Japanese edition (32 / 64bit)
  • CPU: Intel / AMD dual core processor or higher *4 Cores or higher recommended*
  • Memory: 4GB or more * 8GB or more recommended
  • HDD: 1GB or more (400MB or more if the editor is already installed)
  • Display: 1280 x 720 or higher full color
  • Other: Requires Windows compatible sound device for audio playback.
    • Internet connection is required for license activation and update.

Known Problems

  • Version 8.0.9.0
    • Techno-Speech has noted that the timing may be incorrect when a saved project is reopened after adjusting the Timing (TMG) parameter. This issue will be fixed in a future update. Techno-Speech recommends not using timing adjustment parameter until this feature is fixed.[18]
    • In another tweet, Techno-Speech noted that there was another error which in rare cases, if the magnification is any value other than "100%" in the Windows display settings, the Please Wait screen does not appear and the system cannot be started. It was noted a fix was in the works for these bugs[19]
  • Version 8.0.11.0
    • On February 3rd, 2021 Techno-Speech noted that many users have reported errors regarding playback with CeVIO AI and is currently undergoing investigation. Techno-Speech has recommended viewing the CeVIO AI User's Guide page for bug information and troubleshooting.[20]

Gallery

Media Gallery

References

External links

Navigation


Advertisement