Vocaloid Wiki


8,103pages on
this wiki
Add New Page
Comments24 Share


This is an article about the CYBER DIVA software for the VOCALOID4 engine.


The earliest known information related to her dates back to the VY2 release in 2010, in an article a pair of English VY vocals were announced for future development. At the time the developers did not know if they would re-use past samples or go with a new vocal entirely.[1] Cyber Diva was later confirmed to be the female from the pair or "VY3".[2] She was produced in light of research into common complaints towards past pre-V4 English VOCALOIDs. This included the frustrations that came with the mixed British and American phonetic system the VOCALOID engine used for its English VOCALOIDs. It was also discovered that English VOCALOIDs often did not produce the correct sound to match the phonetics in use, caused by accidents in the construction of the VOCALOID either during the recording process or the VOCALOID development process.[3]

Several candidates were tested as possible vocals for the new library. During this process it was noted there were many missing sounds or unusual sounds found they did not expect to be present. There were very clear and expressive sounds present in the constructed libraries. The "Aspiration Problem" that had plagued past English VOCALOIDs was also addressed and fixed. As the problem was fixed, the issue of sounds not matching input symbols were also tweaked. They did not address the issue of expressive tones until they had recorded the right sounds for the right phonetic symbol.


In February, the recording script was rewritten. The aspiration problem was fixed in the new script and the sounds were compacted to make it easier to control the phonetic data better.

In March, two more singers were recorded using the new script. The new script was shorter than the past one and was more difficult for any singer to produce. However, the scripts new results sacrificed expression for clarity and mislabeling was gone. Further tweaks to the scripts were done as recordings went ahead.

Once the voicebanks were brought to a reasonable quality, the team working on her was given the choice between going with the current singers or finding a new one. CYBER DIVA was recorded in 6 pitch layers. She has 3 main pitches, with an additional 3 sets of vowel pitches, aka stationaries. They added 231 different triphones per pitch; more than any other English library released before.

The vocal was tested by two hired English teachers. They were hired was because over half the team were Japanese and sounds that sounded weird to the non-Japanese members did not sound off to them. Due to the mix of British and American sounds, CYBER DIVA did not always sound good when the sounds were mixed. She included the schwa sound; this is not included in all English vocals despite the fact that the phonetic still was used by the voice. A custom dictionary was made for her, this was worked upon in July. Due to its size, they then had to fix the library to make it load faster. The dictionary was finalized in October 2014.


CYBER DIVA was tested for release on the 13th of March.


CYBER DIVA was revealed on the 21st of January 2015[4] as an American accented female produced by YAMAHA and 5 demos followed. Utata-P reported that their demo, "You will See", was done with the default settings and no auto-tuning.[5] She was released on February 4, 2015.[6]

Despite being created with a new Yamaha script, it is worth noting that other English Vocaloids may not be using the same script. As of February 2015, the new script has yet to be shared with Ruby's developer at the very least.[7]

It was also confirmed that knowledge gained from her development was used for Fukase's English vocal.[8]

Starter PackEdit

On October 31, 2016, CYBER DIVA was offered as a starter pack with the VOCALOID4 editor or the VOCALOID4 editor for Cubase. In addition, she was also offered in another set: a starter pack containing her vocal and CYBER SONGMAN along with either the VOCALOID4 editor or VOCALOID4 editor for Cubase.

Product InformationEdit



You can make it big if you want to YouTube
Vocal testings

Mercedes Benz (Growl test) YouTube
(They Long To Be) Close To You YouTube
NAKED EYES (VOCALONET service test) YouTube
It's So Easy (Growl test) YouTube


System requirements

  • Mac OS X/Windows
    • Windows OS:Windows8.1 / Windows8 / Windows 7(32/64bit*)
    • Mac Mac OS X 10.9, 10,8 (32/64bit)
    • CPU:Intel Dual Core CPU
  • Memory: 2GB minimum
  • Hard Drive Space: 4GB minimum

Voicebank LibrariesEdit

Product Information
  Tempo: 60-180BPM  Range: G2-C4
  Trial/Demo Vers?: Yes  Starter Available?: Yes  App vers.?: Yes
Package details
CYBER DIVA is geared towards professional producers. Her voice has an American accent and demonstrates increased clarity over past pre-V4 English VOCALOIDs. She has only one voicebank and cannot use XSY, and is designed for a variety of genres. As a vocal within the VY series of vocaloids, CYBER DIVA is intended to be a product that sets a standard for all other English vocals to follow, being of high and professional quality.

A countepart was later released for her called CYBER SONGMAN.

Vocal traits as noted:
  • Flexible voice
  • Powerful long tones
  • Smooth vocalization
  • Made to be able to sing any genre of music.
  • CrusherP and CircusP both describe her as a "pleasure" to work with.[9]
    • According to the two producers, she has the strength needed for pop ballads.
    • They also say that she can sing heavy words quickly and with clarity.
    • They did not think the "growl" (GWL) feature was meant for her due to being more geared for pop. However, the added harsh, rough timbre makes her able to do rock, soul and blues.[10]
  • According to EmpathP, Cyber Diva is less understandable than Ruby due to her nasally tone.[11]
  • During the interview with the developer on CYBER SONGMAN, it was mentioned that she had troubles sounding like she was actually singing.
Phonetic notes as noted:
  • At 231 triphones, she had the most triphones of any English VOCALOID upon release.
  • According to Prince Syo, she uses the phonetic sound "V" in many words with a reduced vowel when he thinks "I" would be better.[12]
Software issues as noted:
  • As mentioned in her production notes, though she has improved clarity over past English voicebanks, this came at the cost of more expressive tones. Consequently, she may find certain genres which require a softer tone or more emotional singing much more difficult to achieve.
Voicebank sample





Ad blocker interference detected!

Wikia is a free-to-use site that makes money from advertising. We have a modified experience for viewers using ad blockers

Wikia is not accessible if you’ve made further modifications. Remove the custom ad blocker rule(s) and the page will load as expected.