Date: 2021 November 25 21:57
Posted by Joe
The good folks from Japanese software firm Crimson Technology have sent us details about their English language launch of Voidol2. This is real time voice conversion software primarily aimed at streamers, v-tubers and people who have the need to change their voice in real time.
The software has many voice modes and can make you sound like an anime character or even a robot and everything else in between. It retails for 13,200 JPY (about $114 USD / £86 GBP), but is currently on offer for 8,800 JPY (about $76 USD / £57 GBP). It's available to download on Windows or Mac via Sendowl.
Press release as follows:
(Original press release in Japanese: https://crimsontech.jp/rcvoice/voidol2_release/)
Crimson Technology Launches "Voidol2 - Powered by R.C. voice"
A powerful update to the AI real-time voice quality conversion software popular with VTubers and "babi-niku" (virtual female character voice).
In addition to the ultra-high speed new voice conversion engine using synthesizing, the product is also equipped with the sample player function and the function for streaming distribution.
Tokyo, Japan (October 20, 2021) - Crimson Technology Inc. (Headquarters: Setagaya-ku, Tokyo, CEO: Kazuo Hikawa) has released "Voidol2 - Powered by R.C. voice" (hereinafter "Voidol2"), an AI real-time voice conversion software designed for use in content production and distribution applications such as VTubers (YouTube distributors with CG characters).
The Voidol series software has been available for Windows/macOS since 2019. At the time of its release, it was ranked No. 1 in the Mac App Store paid applications, and when the Windows version was released, it was ranked No. 1 in the PC software download store on Amazon.co.jp.
The new product, Voidol2, has been dramatically upgraded to meet the demands of distributors. In addition to the AI real-time voice conversion (AI mode) that is inherited from Voidol1, an ultra high-speed voice conversion engine by synthesizing (SYNTH mode) has been added. With the addition of the sample player function that is indispensable for fun streaming delivery, and the equalizer for detailed sound quality adjustment, you can convert your voice as you wish and deliver it while directing it.
In particular, the Windows version integrates the "Voidol sound virtual driver" that connects to other applications, allowing you to check the converted voice as you deliver.
As a part of our business direction to "market creation for AI entertainment," we developed the industry's first AI voice conversion solution, "R.C. voice (Real Change Voice)" in 2017. Using machine learning, it can change a person's voice into the voice of a specific character in real time. We started marketing of the software from the iOS apps that allowed users to add additional voice models. Currently 22 additional voice models are available to meet the requirements.
Title: Voidol2 - Powered by R.C. voice
Release Date: October 20, 2021
Product URL: https://crimsontech.jp/apps/voidol2/?lang=en
Download URL (through sendowl):
Windows version: https://transactions.sendowl.com/products/78604267/F1907006/view
MacOS version: https://transactions.sendowl.com/products/78604259/62685F9E/view
OS: Windows10 64bit or later
Processor: Intel/AMD Dual Core (Intel Core i5 2.5GHz and above is recommended)
Memory: 4GB or above
Display: WXGA+ (1440x900) or above
OS: macOS High Sierra or later
Processor: Corei5 2.5GHz or above
Memory: 4GB or above
Display: WXGA+ (1440x900) or above
Price: JPY 13,200 (Campaign Price until December 26, 2021: JPY 8,800)
Voidol2 AI Mode
Voidol2 SYNTH Mode
Real-time voice conversion to a specific character (AI mode)
Voidol2 realizes AI real-time voice conversion to a specific character using R.C. voice technology. In addition to the existing Japanese and Chinese models of male and female "Iroha Otomiya" as the preset models, the English female model and the newly developed Japanese female model are released in this application. In addition, from this version, the amount of pitch change can be set, making it possible to make the character reflect the voices of various users.
There are 24 different voice models available for purchase separately.
Newly developed Japanese female model "Urara Otomiya" and English female model "Sophia Lambfield".
A new engine based on synthesizing has been added (SYNTH mode)
The vocoder engine, which has been difficult to operate in real time, has been developed to operate with an ultra-low delay of about 40ms. It is now possible to create a variety of voices, such as pitch changes over three octaves, formant changes, "WHISPER" to make it sound like whispering, and "ROBOT" to make it sound like a robot.
The AI engine and synthesizing engine cannot be used simultaneously in this version.
SYNTH mode parameter screen, DETAIL screen, ANALYZER screen
Click here for the Voidol2 video (with English subtitles)
The industry's first AI real-time voice conversion solution: "R.C. voice"
R.C. voice has been developed by Crimson Technology, Inc. in collaboration with Professor Tomoki Toda of Nagoya (Department of Intelligent Systems, Graduate School of Informatics, Nagoya University) whose research focuses on voice, music and sound environment information processing. "R.C. voice" is the industry's first AI real-time voice conversion solution that can convert a human voice into the voice of a specific character in real time.
The technology has already been used for business applications such as voice quality conversion for show narrators at attractions and voice quality conversion for remote customer service.