Vocaloid Wiki
Advertisement

Template:Infobox Vocaloid

This is an article about the Macne Nana software for the VOCALOID3 engine. This is to be the first VOCALOID release of the Macne Series.

History

In 2013, VOCALOID NEO (a version of VOCALOID3 for the Mac) was released. However, though the news was interesting, as a Magazine editor Haruna Ikezawa had no budget for a VOCALOID production. Later, after things were sorted to produce the VOCALOID3 version, teasers were left hinting at Macne Nana being produced for VOCALOID3. This was later confirmed on Macne Nana's Twitter account.[1][2]

According to Mac Fan magazine, there was a conversation with Yamaha; no money was involved and the OK was given for the voice of Haruna Ikezawa to become a VOCALOID. The vocal would be reviewed at 70% completion. If the vocal does not meet with expectations, it will be scrapped. It will only be put forward to commercial release if the vocal is satisfactory.[3]

Because of the lack of a budget, a sponsor had to be found to power the projects budget and the distributor had to be sorted also. In the end the issue was sorted with Bplats. The decision was made to make this a Vocaloid Store exclusive only. Originally the project was to be released in Jan, but Haruna Ikezawa wanted to challenge herself by doing a English vocal as well as she found the original recording of her vocal boring. Permission, however, had to be asked due to the time frame that had been set aside to produce the vocal. However, the go ahead was given and English went also into production. Adding English set the project back, making it take a years and a half to produce and delayed release.

During the summer, her Japanese and English was re-recorded to improve its vocal. The Japanese vocal took a day to record while 5 days were taken to record the English again. A Chinese vocal was also produced as Haruna Ikezawa knows Chinese, but production was stopped at a prototype stage. This vocal was left to be considered as a future release rather than an immediate one.

Editing began in August, of which much trial and error occurred. Due to the Jan deadline, only 3 months were left to edit the vocal. The vocal had to be delivered to Yamaha by November. A total of over 1,000 hours were put into producing the vocal, this amounted to a workload of 15 hours a day. The Japanese vocal had to be redone 6 times, the English about 11 times. The English vocal in particular was difficult for Haruna Ikezawa to work with.

In October 2013, it was confirmed she was going ahead and that she would have both a PC version and a Mac version. It was also confirmed she would include an English as well as a Japanese vocal. She was slated for a January 2014 release.[4][5]

In November 2013, it was revealed that Bplats, Inc. will be the company that records her voicebanks. It was also stated that her English and Japanese voicebanks can be bought separately or as part of a bundle. The bundle went on pre-order in mid-december 2013. Those who pre-ordered would receive several bonuses, including the Macne Nana 2S vocal and wallpapers. Those who pre-ordered were also entered into a competition to win one of 10 signed hand-drawn pictures by Haruna Ikezawa and Gomoku Akatsuki.[6]

Future development

It was mentioned in a interview that the Chinese vocal was not completely out of question yet, as they made a test vocal for Nana but it is not a complete version, they don't guarantee release for it.[7] Haruna Ikezawa was also interested in the French Vocaloid by VoxWave which was being produced with the help of VocaTone. Since Haruna Ikezawa can also speak French, she put the idea of French Nana also on the table for future developments.[8]

Currently the French project has been brought to a halt, decreasing Macne Nana's chances of getting a French vocal for the time being.[9]

AHS Sales

In May 2015, due to the closure of the Vocaloid store, there was talk of selling Macne Nana in the AH-Software Co. Ltd. store. Although the VOCALOID SHOP was replacing the store, at the time the clause was considered "messy" – there were too many uncertain factors regarding what would happen after the switch.[10] As of June 2015, Macne Nana officially became available from the AHS shop.[11]

Retirement

On March 28, 2016, Macne Nana Plus was announced and was revealed to be the VOCALOID4 update for Nana. The package was to include an improved version of Nana's VOCALOID3 voicebanks and also added Macne Petit. However, it was also confirmed that Nana's VOCALOID3 voicebanks would cease distribution on March 31.[12]

Product Information

Demonstrations

Demonstrations

ひいらぎ飾ろう (Hiiragi kazarou)/(Deck the Halls) (Japanese and English) SoundCloud
さくらさくら (Sakura Sakura) (Japanese) SoundCloud
Twinkle Twinkle Little Star SoundCloud
Na☆な! マクネナナ (Na☆na! Macne Nana) (English and Japanese)
SoundCloud
Na☆na! Macne Nana (Original Garage Vb) for comparison purposes

Recommended

TBA

Voicebank Libraries

GENERAL USAGE

Product Information
Optimum
  Optimum Range: C#3 ~ C#4  No. of Keys: W ~ 7, B ~ 6, Total ~ 13
Versions
  Trial/Demo Vers?: No  Starter Available?: No
Package details as noted:

This is the VOCALOID version of Macne Nana, which contains both Japanese and English vocals. Both a PC and Mac (VOCALOID NEO) version are available for sale. Her VOCALOID vocal maintains the same intentions as the vocals built for the original Reason and Garageband vocals; she is meant to sound light, cheerful, youthful and cute. The Japanese and English Macne Nana banks can be bought either separately or as a bundle.

The Macne Nana 2S vocal and wallpapers were included in pre-orders of the package.

Vocal traits as noted:
  • This is the very first time a Macne vocal has been produced with English capabilities. Previously, all versions of the Macne series and its vocals possessed only Japanese capabilities.
  • Vocals are much smoother than her original version; overall quality is much higher and, as a result, she produces a less robotic sounding voice.
  • Haruna Ikezawa is noted to be a fairly adapt singer but is still not a professional vocalist, so Macne Nana's vocal quality is also on par with that of a amateur vocalist.
  • According to the YAMAHA store, Macne Nana is best at pop songs.
  • She was edited to meet a 24bit/96kHz quality mark. This was achieved via wavelength of the recordings of the vocal being edited.
  • The equipment was carefully selected during the recordings to minimize loss of characteristics of her provider's tone.
Software issues as noted:
  • Users of previous Nana vocals will note the differences between the VOCALOID version and past versions. This was notable even in the original importation of the vocals from Garangeband and Reason to UTAU. This is owed both to the differences of engine workings as well as the vocal performance of Ikezawa's voice acting across the different versions that were recorded of her vocal.
    • The original voicebank itself often ended up quite high-pitched in comparison to the VOCALOID version when used for its intended software. This is because it was comparable to a sample pack with a list of the needed sounds and didn't have have a fully adapted vocal synthesizer attached to simulate the human voice correctly and played the samples in their raw state.
    • When it was imported into UTAU, the original Nana vocal then became lower in tone compared to the VOCALOID version and at times much lower then when used in Reason and Garage Band.
    • Prior to this, the UTAU 2S voicebank included in her 2S release is considered the most correct version of Nana's voice. It is also the one the VOCALOID voicebank most closely resembles.
Cross-Synthesis as noted:

Macne Nana Japanese remains unable to XSY and once imported into the VOCALOID4 will not register for XSY at all.

Once imported into VOCALOID4, her English voicebank becomes available for XSY. However, it has no additional Nana English voicebanks it can collaborate within the release itself. Users will have to purchase further voicebanks from other released within the XSY group "AHS English".

  • In June 2017 a new XSY group for English was created for AHS Software vocals. All vocals within this group are compactiable with each other for XSY purposes. This vocal release is part of this group. If a User owns one or more vocal releases within the "AHS English" group, XSY between them will open up.
    • In total this group offers the equivalent of 2 additional voicebanks achievable via XSY. The total tone variation offered by the group comes to a theoretical 4 voicebanks in total.
    • This is currently the smallest of the multiple character XSY groups, with just 1 character in it currently; Macne Nana
    • There are currently 2 unique voicebanks within this group; "Macne Nana English" and "Macne Nana English V4"
  • Note that the addition of this group did not open up GWL for Macne Nana VOCALOID3.

Individual Voices

Product Information
Optimum
  Optimum Range: C3 ~ C4  Optimum Tempo: 77 ~ 177 BPM
  Total Tempo (min-max): 100 BPM  No. of Keys: W ~ 8, B ~ 5, Total ~ 13
Package details as noted:

This is the Japanese Macne Nana vocal.

Vocal traits as noted:
  • The quality of this vocal is much higher than all past versions built for Nana.
  • It is also "correct" for her character and original design, which is something the UTAU vocals struggled to maintain, having often being lower in pitch or tone.


Voicebank sample

Macne Nana Japanese VOCALOID3


Product Information
Optimum
  Optimum Range: C3 ~ C4  Optimum Tempo: 77 ~ 177 BPM
  Total Tempo (min-max): 100 BPM  No. of Keys: W ~ 8, B ~ 5, Total ~ 13
Package details as noted:

This is the English version vocal. She has a strong Japanese accent due to the choice being made during development to let the vocal sound natural over sounding accurate.

Vocal traits as noted:
  • Her vocal has a good level of clarity despite the Japanese accent.
  • Due to having the same tempo, range and tone, switching mid-song to the Japanese vocal is less noticeable.
  • Nana has some of the smoothest English sounds overall in the billingual voicebank releases and transistions from sound to sound naturally. This is notable as other billingual voicebanks such as Hatsune Miku V3 English have issues with how unnatural their sounds are overall.
Phonetic notes as noted:
  • She has trouble pronouncing her R sound properly, it can be mistaken for an L or W sound.
  • A common misconception is that Nana lacks vowel sounds amd is even missing some, but unlike some cases such as Megurine Luka's original vocal VOCALOID2 vocal, this is not the case and she actually has all of the recorded vowel sounds. She has one of the strongest issues with a lack of distinction in her vowels meaning, though she has all the needed vowels, they at times sound so similar that the differences aren't noticed much. This can be a limiter when it comes to using her vowels in word construction, especially for non-English words.
Software issues as noted:
  • Nana English has a very strong Japanese accent, with it being arguable if it is or isn't the strongest. Nana has the most natural sounding of the billingual voicebanks. This leads to producers at times having some misconceptions on the voicebank and Nana may need special considerations when using her. It is unwise to presume techniques that work for other billingual voicebanks will work for Nana's.
    • In some cases, it could even be better not to try and attempt to fix the lack of distinction in her voice.
Voicebank sample

Macne Nana English VOCALOID3



References

Navigation

Template:Nav MI7Template:Nav bplatsco

Advertisement