Silhouettes for SF-A2 miki, along with the other VOCALOIDs by AHS, were revealed on a poster on November 6, 2014, however, at the time, it was unconfirmed what they were for. On November 20, it was confirmed in the livestream for VOCALOID4 that they all, with the exception of Tohoku Zunko, will be receiving VOCALOID4 voicebanks.
On May 14, 2015, AH-Software confirmed that miki would only have a Natural voicebank and would release on June 18, 2015. It is possible to buy her Natural voicebank, but there was also information about a starter pack. Other details such as tempo, range, the prices for her packages were also revealed. It was also shown that miki V4 has an official MMD model by KIO.
On June 4, a trial version for miki V4 was made available.
As her original package, the performance while singing is soft in lower notes and progressively gets powerful and sharp in higher. While the transition of both singing styles was unstable in the original library, it was improved to make it more natural and smoother.
Improved and more accurate pronunciation and better audio quality.
Most of the glitches from the V2 vocal were fixed, making the new voicebank more user friendly.
She becomes softer in higher notes and has less tension than her V2, which doesn't allow her to work well as her V2 in more fast-paced music.
In contrast, her stationaries became stable enough to maintain a good performance in slower and mellow genres.
She will register for use of XSY though naturally does not have an accompanying vocal for XSY.
As of Ver.4.3.0 of the VOCALOID4 engine, AH-Software Co. Ltd. have opened up XSY to work with all their VOCALOID3 and VOCALOID4 vocals except Kaai Yuki V4, so if a User owns another of their vocals XSY between them will open up.
In total if user was to own every VOCALOID3 and VOCALOID4 Ah software vocal, this offers the equivalent of 156 additional voicebanks achievable via XSY. The total tone variation offered by the combined packages comes to a theoretical 169 voicebanks in total.
Phonetic notes as noted:
As her V2, she will still skip the "-" phoneme.
Software issues as noted:
The new voice uses the same recordings Miki Furukawa did for SF-A2 miki V2, so aside from the quality and triphonetic data included, there are no really important or noticeable changes.
Compared to the original, this vocal sounds more sharp and isn't always smooth as a result.
miki V4 has the same problem as her V2, she will often sounds off-key, though this issue is now easier to fix with minor editing.