Vocal Synthesizer Wiki
Advertisement
šŸ’» Technology article work in progress. What is being worked on: History, Companies involved in the development
For information on how to help, see the guidelines.  More subjects categorized here.
šŸ’»

VOICEPEAK is a text-to-speech synthesizer developed by Dreamtonics Co., Ltd. in collaboration with AH-Software Co. Ltd.. VOICEPEAK is equipped with Dreamtonics' AI speech synthesis Syllaflow technology that can be used to easily read aloud words and sentences input via text using high-quality voice synthesis.

History[]

On September 18, 2021, Dreamtonics revealed that they were working on a text-to-speech synthesizer that was described to be "ultra-realistic".[1]

On February 17, 2022, this was revealed to be "VOICEPEAK" (also stylized as "voicepeak" or "VOICE PEAK"), which was developed in collaboration with AH-Software.[2] According to AH-Software's product page, VOICEPEAK was equipped with the high-quality AI speech synthesis engine "Syllaflow", which was developed by Dreamtonics.[3] A trial version of VOICEPEAK, along with one male and one female voice, was released on the same day, with the commercial version set to release for Windows, MacOS, and Linux on March 11 with four females and three males.[4]

On August 11, Dreamtonics teased that new Chinese speech voices were expected to debut the next day, but did not state the engine.[5] On August 12, this was confirmed to be VOICEPEAK and four voices were shown: an adult female, an adult male, a middle-aged or elderly male, and a young female. They were shown to speak with different moods and tones. A release date had not been announced, but more demos were to be expected.[6]

On November 18, AH-Software announced a new character-themed lineup of voice libraries for VOICEPEAK, including Frimomen, Asumi Shuo and Ririse, and Tohoku Zunko (the latter also bundled with Zundamon as a bonus); they were originally expected to be released on December 15,[7] but staff illness delayed the release until January 13, 2023.[8][9] These VOICEPEAK libraries are being marketed in the same fashion as VOICEROID libraries and thus can be seen as the spiritual successors to VOICEROID (AHS' old packaging of AI, Inc.'s AITalk speech-synthesis engine).

Demonstrations[]

Demonstrations

6 Narrator set + 1 demo YouTube
Tohoku Zunko demo YouTube
Asumi Shou demo YouTube
Asumi Ririse demo YouTube
Frimomen demo YouTube
Additional Narrators demo YouTube
Jashin-chan demo YouTube
Tohoku Kiritan demo YouTube
Koharu Rikka demo YouTube
Haruno Sora demo YouTube
Otomachi Una demo YouTube
Kyomachi Seika demo YouTube
Tohoku Itako demo YouTube
Miyamai Moca demo YouTube
Tsurumaki Maki demo YouTube

Requirements[]

  • OS:
    • Windows: 11/10/8 or later
    • Mac: OS X 10.11 or later
    • Ubuntu: 20.04 or later
    • Raspberry Pi (Modelļ¼šRaspberry Pi 4 / OSļ¼šRaspberry Pi OS (64-bit))
  • CPU:
    • Windows/Intel Mac/Linux: Intel Core i3 or higher or equal AMD processor
    • Apple Silicon Mac: Apple M1 or higher
  • Memory: 2GB
  • Storage: At least 500MB of available space for installation (SSD recommended)
  • Display Resolution: 1920x1080 or higher recommended (1280x720 required)
  • Others: DVD-ROM drive (packaged version), audio device, Internet connection required

Releases[]

Commercially Viable Product Series

Character Product Series

Trial releases


Upcoming voicebanks[]


Discontinued voicebanks[]

Official Derivatives[]

Official characters licensed by AH-Software or its affiliates that use or are intended to be used with VOICEPEAK vocals.[10][11]

References[]

External links[]

Official[]

Articles[]

Navigation[]

Advertisement