On January 20, 2011, confirmation was released that recordings had started for the V2 Megpoid Extend, several vocal recordings having already been recorded.
The first voicebank demo, Megpoid Extend Power β was uploaded on May 21, 2011 to Nico Nico Douga, however its time spent up was only brief. Demos for Power and Whisper α were also put up, but also removed a short time later. A more permanent version was uploaded later. GUMI's whisper Extend was used to sing "Fly me to the moon" and "BeMyself" in her power Extend; their uploads were also temporary and were removed at a later date.
By June it was mentioned that there was 4 vocals in the extend development and they were being pushed to enter β by 26 of Jun.
It was mentioned in September that they had performed checks on 1,200 connections for V3 Megpoid, this was done 5-6 times for roughly every 1GB of data. Modifications occurred where necessary.
The Extend idea was later dropped and instead they were launched as new voicebanks for VOCALOID3. According to Noboru, "Extend" was dropped from the name because it would cause confusion on whether or not VOCALOID2 Megpoid was needed to use the vocals. The voicebanks are sold individually or as a pack. The new update of information confirmed a October 21 release date. The term "Extend" was dropped from the product name entirely.
Megpoid Talk is a speech software version to allow GUMI to do narration was discussed in October 2012 on Noboru's twitter account, but no details were given at the time. However, there was later a demo on December 24, 2013, and a release date set for January 16, 2014.
This is a speech software which will allow GUMI to talk for the first time. It allows delicate editing of her vocals. It is powered by the engine "FineSpeech Ver 3".
Originally intended as a expansion pack for Megpoid, V3 Megpoid was released as a separate Megpoid vocal. They are sold as a "complete" package (excluding the V3 Megpoid - Native vocal) or separately on their own. These vocals are natural sounding and offer a simple change of tone, resulting in V3 Megpoid keeping professional singing results. GUMI's new voicebanks do not require Megpoid or V3 Megpoid - Native software to use.
Vocal traits as noted:
All vocals share the same optimum range and tempo making changing mid-song relatively smooth.
Since they are a simple change of tone, they are known to sound close to the original Megpoid at times.
Jointly, they address the main issue of the Megpoid vocal, in that is was highly resistance to editing change. This limited the number of tones it could produce in comparison to other VOCALOID2 vocals.
These 4 vocals also do so without sounding too unnatural or different in the process, an issue that the VOCALOID2 "Append" releases faced.
When any vocal from this package is imported into VOCALOID4, the cross-synthesis (XSY) function can be used if at least 1 other vocal from this package has also been imported.
In total the package offers the equivalent of 12 additional voicebanks achievable via XSY. The total tone variation offered by the package comes to a theoretical 16 voicebanks in total.
In addition, V3 Megpoid - Native also can be cross-synthesized with any vocal from this package if it is also imported.
Once imported, all vocals offer the equivalent of 20 additional voicebanks achievable via XSY. The total tone variation offered by the package comes to a theoretical 25 voicebanks in total.
From Ver.4.3.0 of the Vocaloid engine, a XSY group "Internet" was added to Vocaloid. All vocals within the "Internet" group can XSY with each other. This vocal release is part of this group. If a User owns one or more vocals within the "Internet" group, XSY between them will open up as a result of owning this release.
The "Internet" group is currently the largest XSY group. In total if user was to own every VOCALOID3 and VOCALOID4 XSY capable INTERNET co., LTD vocal, this offers the equivalent of 812 additional voicebanks achievable via XSY. The total tone variation offered by the combined packages comes to a theoretical 841 voicebanks in total.
Note that Users of the updated XSY list have to understand that as confirmed by Internet co., Ltd themselves, depending on the combination of the two vocals a level of extra unwanted noise may be generated.
Another note worthy detail on V4 and later vocals compared to V3 vocals within this group is that Internet Co., Ltd. increased the number of triphones from Chika onwards. Thus, many newer vocals do not match sound-for-sound with the older pre-Chika released vocals.
Post VOCALOID4 releases were also developed with XSY in mind, so are better for XSY then the older VOCALOID3 releases.
In addition to not being built with XSY in mind, VOCALOID3 packages with only 1 vocal in their contents that were releaseddmay be even more problematic for XSY. Internet co often made multi-pack releases have a quality that made switching mid-song less noticeable, making them slightly better for XSY then single vocal releases.
Note that also "GWL" does not become viable for use with any V3 vocals used as the primarily vocal. This should be considered when using any vocal for XSY within the "Internet" group from V3.
The software is also capable of being used as a "control" for XSY in a secondary role. It is possible to switch out the vocals to get a slightly different result on a primary vocal giving the vocals the ability to act as a "utility" purpose in XSY. This is due to the vocals having the same range and tempo between them, so if one vocal does not produce a desired result aother with a slightly different result can be easily switched in. However, due to having double the vocals, the later Megpoid V4 is a better option for this purpose.
Software issues as noted:
Because all V3 Megpoid voices have the same tempo and range, there is no major advantages between the Completed voices or Native when compared to other packages such as Hatsune Miku V3 or Meiko V3.
All Megpoid V3 vocals had issues related to leading in vocals causing a noise.