quinta-feira, 4 de março de 2010

VOCALOID-flex: A New Synthesis Engine


The Yamaha Corporation is developing a new form of software synthesis engine, the "VOCALOID-flex". This features not only the traditional VOCALOID singing synthesis, but it includes "talking" that can be achieved by adding additional sound files.
The general public is enthusiastic with the ability to add further expression to the joy of growing voice synthesis by the user with the NetVOCALOID (announced in April 2009). In combination with VOCALOID, it has a wide range of new users with the possibility of a widespread penetration goal.

"Background of Development"
VOCALOID works by simply typing the lyrics and setting the melody. Their songs and YouTube videos from Nikoniko the announcement of video portals such as birth and various topics in the news as often popular as a social phenomenon called were well.
Under these circumstances, the recent dialogue and screams in songs such as "speaking" the place of using the VOCALOID software that comes with music production and editing products, post videos to create a portal site data within a limited range. Some users have started to appear, so far is the realization of VOCALOID had been targeted for speaking to singing only on having a lot of effort and time, "talking". There was a limit to the unique power of expression.
Also, the mainstream of conventional speech Text to Speech (less TTS) Will they be able to synthesize a rich voice and representation of human emotion and dialect is difficult to express the richness of human emotion than new was hoped the speech synthesis system.
We are based in this context, TTS and speech synthesis in different forms such as, original VOCALOID rich synthesis technology utilizing the power of expression in the form of singing with the "talking" to achieve a developed has been.


VOCALOID-flex is newly developed, subtle changes are required compared to the voice sound "talk" in order to achieve the impossible with conventional VOCALOID, the phonological (sound configuration and length and phoneme) and prosody (pitch, intensity) and speech synthesis engine is software that enables voice and minor editing.
In concrete terms, which could not vowel devoicing (pronounced as voiceless vowels cases (eg: North [ki ta] of [i])) and of omission (not case the vowel (eg: - I [de su] the [u]) can be expressed, can now be edited over the height and strength and the length of the consonant sounds. As a result, the speech can be so close to humans from and nuances in the voices, and can now also put the accent and intonation and expressive dialect.
However, VOCALOID-flex library for use with audio material without adding a new sound a new voice will be possible to utilize existing libraries.


Our future, VOCALOID-flex in the business for each company are interested in using speech, we will gradually begin offering. Based methods will provide the software license, other companies have already deployed speech business, is planning to offer such a wide range of entertainment across all industries. There are no plans to provide the consumer.
VOCALOID-flex based, the licensee company, including the traditional PC business as VOCALOID software, and Internet business, built into the hardware, can be used widely. A view to expand our overseas VOCALOID-flex is also compatible with English. VOCALOID-flex licensee companies using the "talk" to achieve the service is to provide software synthesis engine from Yamaha VOCALOID-flex and Yuzaintafesuapurikeshon VOCALOID-flex is the need for control.
Yuzaintafeisuapurikeshon and is specific to the "talking" to edit the application to achieve the phonology and prosody, but we will support Kimasu available on their own licensees.
As a concrete example of the future business development, entertainment areas are expected to use in applications such as dubbing for animated films and video content. Also, you can bring out the robots, and probably also to the development of character toys. Other areas are valid and used as an emergency alert voice and workshop.
VOCALOID-flex business development utilizing one of the bullets, the most popular game, Konami Digital Entertainment Co., April 29 will be released on "METAL GEAR" series latest "METAL GEAR SOLID PEACE WALKER" (Kana notation: Metal Gear Solid Pisuuoka) is. We decided to use the voice of the character creation tools of this appeared.
[Note 1] "Hatsune Miku" is Crypton Future Media (shares) registered trademark. Voice "Saki Fujita," is being put together the original vocal voice sound cute characters, played by Mr. Pop.
[Note 2] "etc. like study of" The (stock) Internet trademark. GACKT singer being put together with the original vocal sound of the voice, GACKT can still get the sound with the voice quality of voice synthesis.
Product name in the text, and name are trademarks or registered trademarks of their respective owners and our company.

Sem comentários:

Enviar um comentário