VOCALOID Hatsune Miku VOCALOID6

[MagicalMiku]

Hatsune Miku VOCALOID6
A new version of "Hatsune Miku V6 AI" is scheduled to be released within the year!

Original artwork by iXima made for Miku 16th Birthday (August 31st, 2023 via twitter)
Crypton is currently planning to release a new version of the singing voice synthesis software "Hatsune Miku V6 AI" which uses Yamaha Corporation's VOCALOID6 as a new singing voice synthesis software using AI technology, which has long been requested by creators.
For "Hatsune Miku V6 AI", which is currently under development, we will set a preferential price for regular users of related software from "Hatsune Miku V4X" onwards. Details of each will be announced at a later date, so please wait for further information.
press release:

バーチャルシンガー6名の歌声をまとめた新製品の予約を開始！　「初音ミクV6 AI」も年内リリースに向けて準備中！

クリプトン・フューチャー・メディア株式会社のプレスリリース（2024年8月1日 14時06分）バーチャルシンガー6名の歌声をまとめた新製品の予約を開始！　「初音ミクV6 AI」も年内リリースに向けて準備中！

prtimes.jp

(note: this press release was written on August 1st 2024, together with the announcement of Hatsune Miku NT ver.2 and Piapro Characters Super Pack.)

Development Progress:
- On December 26th 2024, Takamichi Koizumi of Crypton Future Media wrote an Apology public message on Sonicwire official blog:

Thank you very much for your continued patronage of Crypton Future Media products.
As announced in our August 1, 2024 press release, the virtual singer software "Hatsune Miku V6 AI" currently under development was scheduled for release in late 2024, but due to development reasons, the release has been postponed. As we continue with development, we have come across some issues that are difficult to overcome, and after consulting with YAMAHA, we have concluded that it will take some time before we can deliver stable singing that is characteristic of Miku.
We sincerely apologize for the inconvenience and concern that this postponement of the release has caused to everyone who was looking forward to the new product. The planning and development team would like to offer our sincerest apologies.
Crypton Future Media, Inc. "Hatsune Miku V6 AI" planning and development team

- During the spring of 2025, both Yamaha and Wataru Sasaki of Crypton confirmed that the voice quality of Miku V6 reached its climax, and they are "making steady progress" (via twitter)

- Crypton Future Media will celebrate its 30th anniversary on July 21, 2025. There is currently a 30% discount on all Piapro characters until July 31st, and it's possible there will be some news about Miku V6 during the 30th Anniversary week.

Announcement
Hokkaido University, which has a connection to us, will be holding a launch event for our company CEO Ito's book, "The Future of Creation: Why Hatsune Miku Was Born in Hokkaido"!
Non-students are also welcome to participate, so if you live near Sapporo, please check out the details

<Hokkaido University x Hokkaido University Co-op>
To celebrate the release of "The Future of Creation"!
"The creator of Hatsune Miku" Ito Hiroyuki talk and autograph session!

■Date and time: July 29th (Tue) 17:00~
■Venue: Hokkaido University Frontier Applied Science Research Building, 2nd floor (via twitter)

- On August 31st, 2025 for Miku birthday, Crypton staff will hold a talking session at Magical Mirai 2025 in Tokyo:

Virtual Singer Product Stage
Date and Time: Sunday, August 31, 2025, 11:00-11:45
Stage Introduction: We will look back on the Virtual Singer products that Crypton Future Media has released so far and talk about future developments.
* This stage will be live-streamed and archived on Crypton's official YouTube channel, "Piapro TV."

[TOKYO]企画展ステージ｜初音ミク「マジカルミライ 2025」

初音ミク「マジカルミライ 2025」は、創作で繋がる初音ミクたちのライブ＆企画展です！【SENDAI】8/1(金)-8/3(日)、【OSAKA】8/9(土)-8/11(月・祝)、【TOKYO】8/29(金)-8/31(日)開催！

magicalmirai.com

- Crypton 30th Anniversary special interview - Episode 2 "Music x Technology":
https://vocaverse.network/threads/crypton-future-media-30th-anniversary-special-interview.10180/post-59353

--By the way, I'm sure there are people reading this who are wondering, so I'll ask you... How is the progress of "Hatsune Miku V6 AI" going?

Sasaki : Regarding the "challenges that are currently difficult to overcome as we proceed with development," which we reported on the SONICWIRE blog in December 2024, we are now seeing a clue to solving them with the cooperation of YAMAHA. We are proceeding with development so that we can make it possible for you to try the new features in some form by 2025, so we apologize to everyone who is looking forward to it, but we would appreciate it if you could wait for a little while longer.

about the future development of Rin, Len, Luka, Meiko and Kaito:

Ito : When it comes to new initiatives such as "NT" and "V6 AI," "Hatsune Miku" always takes the lead, but of course we are also thinking about other virtual singers. It's taking time, but we hope you will continue to watch over the developments going forward.

official website:
https://sonicwire.com/product/virtualsinger/special/index
Crypton official page:
https://ec.crypton.co.jp/pages/prod/virtualsinger/
Amazon JP product page:
coming soon!
support page:
https://sonicwire.com/support/prod

previous versions:
https://sonicwire.com/product/virtualsinger/special/superpack
https://sonicwire.com/product/virtualsinger/special/mikuv4x
https://sonicwire.com/product/virtualsinger/special/mikuv4c
https://sonicwire.com/product/virtualsinger/special/mikuv3
https://ec.crypton.co.jp/pages/prod/virtualsinger/cv01a
https://ec.crypton.co.jp/pages/prod/virtualsinger/cv01

 

[Palamon]

Coming soon, April 31st, 2026. /j

 

[InstallGentoo]

Coming soon, April 31st, 2026. /j

with all the computer merch collab with asus it feels like that's the most affordable release date for all the diehard fans (the ones with very little impulse control, like me).

 

[kimiworld]

Coming soon, April 31st, 2026. /j

If and only if it is confirmed to be April 31st 2026, at least I would have time to be prepared (money wise) lol.

 

[Bookworm2]

Coming soon, April 31st, 2026. /j

2026? No, minimum 2028.
That said, I am excited for Miku V6. Who knows, maybe this means updates for the rest of the cryptonloids! I wonder if the tone will be similar to a previous voicebank, or if it will be unique. Or is there an option for A.I. tracks similar to appends? I really don't know much about V6. Either way, I look forward to hearing it!

 

[MagicalMiku]

2026? No, minimum 2028.
That said, I am excited for Miku V6. Who knows, maybe this means updates for the rest of the cryptonloids! I wonder if the tone will be similar to a previous voicebank, or if it will be unique. Or is there an option for A.I. tracks similar to appends? I really don't know much about V6. Either way, I look forward to hearing it!

The main issue of Vocaloid6 is that, while it's a powerful engine with lots of interesting features, the change of tone is so much different compared to Vocaloid 3 and Vocaloid 4, so there are some issues that need to be solved in order to get a good tone that can be compared with the previous Vocaloid versions.
It has already been hinted last year that also Rin, Len, Luka, Meiko and Kaito might get Vocaloid 6 (together with NT2), but for now the focus is on Miku Vocaloid 6, because then it'll be easier to start from a good base for any future development
Those issues are the main reason why some years ago Crypton started development on NT, and I think Vocaloid 6 will especially be used for english and chinese voice banks of Miku and other Piapro characters. It's good that the producers will have more choices, and it all depends on what kind of "sound signature" they are aiming for. For example, some producers, even the most popular ones, still use today Miku V2 and her Append<3

 

[InstallGentoo]

some producers, even the most popular ones, still use today Miku V2 and her Append<3

That makes sense, V2 and append are legendary voicebanks. V2 is the voicebank of Miku in the popular section, and appends I recognize it as the cult classic that held up the trance and edm side of the fandom.
These days her NT banks are competing with synthV and Teto and it apears to me Teto is more popular due to it being relatively newer and shinier. Or maybe it's percived and I just been seeing more of her than I personally would like to see (no offense to the teto fans). So in terms of "cutting edge" and "future" perceptions of the public I think CFM is in the undertow in this race for the general public at the moment. V6 would be the answer to that I think, so it would make sense for them to try and get it just right when it releases.

 

[sunnyp4rk]

I think the thing I'm most interested in is how the XLS for English and Chinese will sound for her. V6 is already iffy on quality for those languages, depending on the voicebank. I want to see if it really is feasible for Miku to have only one voicebank that can sing in all three languages, rather than 3 different products.

 

[JikyuVox]

I think the thing I'm most interested in is how the XLS for English and Chinese will sound for her. V6 is already iffy on quality for those languages, depending on the voicebank. I want to see if it really is feasible for Miku to have only one voicebank that can sing in all three languages, rather than 3 different products.

Agreed! That's why I want Miku V6.

 

[Blue Of Mind]

some producers, even the most popular ones, still use today Miku V2 and her Append<3

That makes sense, V2 and append are legendary voicebanks. V2 is the voicebank of Miku in the popular section, and appends I recognize it as the cult classic that held up the trance and edm side of the fandom.

These comments have just made me think if Crypton had just kept reissuing V2 Miku in some way (and even V1 Meiko and Kaito and V2 Rin, Len and Luka), a lot of fans would have been less upset over the years. With every engine upgrade and attempt to keep their V1 and V2 tones, ironically the Cryptonloids sound less like their original banks. (This is my biggest issue with Miku NT - it tries to sound like V2 Miku but it's off. It's hard to explain to newer fans, but if you've been listening to Miku for so long like I have, you can tell immediately.) Alas, Yamaha has long stopped supporting V2, and leaving such an ancient program up for sale would increase compatibility issues with newer OSs. (How do V1 and V2 work on Windows 11, anyway?)

 

[PearlStarLight5]

(How do V1 and V2 work on Windows 11, anyway?)

I don't know about V1, but I'm able to use V2 on Windows 11

 

[sunnyp4rk]

These comments have just made me think if Crypton had just kept reissuing V2 Miku in some way (and even V1 Meiko and Kaito and V2 Rin, Len and Luka), a lot of fans would have been less upset over the years. With every engine upgrade and attempt to keep their V1 and V2 tones, ironically the Cryptonloids sound less like their original banks. (This is my biggest issue with Miku NT - it tries to sound like V2 Miku but it's off. It's hard to explain to newer fans, but if you've been listening to Miku for so long like I have, you can tell immediately.) Alas, Yamaha has long stopped supporting V2, and leaving such an ancient program up for sale would increase compatibility issues with newer OSs. (How do V1 and V2 work on Windows 11, anyway?)

If they did what other companies did to port older voicebanks to newer versions of Vocaloid (Gumi V2 -> V3 through her Native vb, any V3 that simply had growl added to their voicebank when ported to V4 without rerecording, etc.), I would be fine with that (maybe this is just me wanting an easier way to buy Luka V2 lol).

NT's biggest issue IMO is that it's choppy, while any Vocaloid version of Miku hasn't been affected like that before. I'd love to know WHY NT is so choppy. I'd love to also know why they think it's ok to release a product of that quality (I know we've had Gumi English before, but English is a more difficult language to make a concat vb for).

 

[Bookworm2]

(How do V1 and V2 work on Windows 11, anyway?)

From what I understand by reading tutorials, it plays ok with win 11, and actually gets a bug fixed if run on 11. I don't have any V1 voicebanks though, so I can't tell that much.

 

[Blue Of Mind]

If they did what other companies did to port older voicebanks to newer versions of Vocaloid (Gumi V2 -> V3 through her Native vb, any V3 that simply had growl added to their voicebank when ported to V4 without rerecording, etc.), I would be fine with that (maybe this is just me wanting an easier way to buy Luka V2 lol).

NT's biggest issue IMO is that it's choppy, while any Vocaloid version of Miku hasn't been affected like that before. I'd love to know WHY NT is so choppy. I'd love to also know why they think it's ok to release a product of that quality (I know we've had Gumi English before, but English is a more difficult language to make a concat vb for).

Admittedly, Miku V2 had choppy transitions as well, but people forgave it in hindsight because she was the second V2 ever released after Sweet Ann, and the very first Japanese V2. Nowadays, some people find V2's choppiness charming, but obviously that's nostalgia over older software. I agree that a modern software like NT shouldn't be having choppy results. At least Gumi English V3's choppiness can be manually fixed within Vocaloid if you look up tutorials (though apparently, it's time-consuming to do so).

From what I understand by reading tutorials, it plays ok with win 11, and actually gets a bug fixed if run on 11. I don't have any V1 voicebanks though, so I can't tell that much.

Ohhh, that's good! I've heard some horror stories in the past about V1 crashing on Win7 and 10, so I was curious how V2 was running on other modern OSs.

 

[sunnyp4rk]

Admittedly, Miku V2 had choppy transitions as well, but people forgave it in hindsight because she was the second V2 ever released after Sweet Ann, and the very first Japanese V2. Nowadays, some people find V2's choppiness charming, but obviously that's nostalgia over older software. I agree that a modern software like NT shouldn't be having choppy results. At least Gumi English V3's choppiness can be manually fixed within Vocaloid if you look up tutorials (though apparently, it's time-consuming to do so).

Ah, in my defense, I've never played around with V2 Miku lol. I guess it can be excused since she DID come out in 2007 (almost 2 decades ago), so older software having issues is a normal thing. Like you said, modern software like NT should not be as choppy as it currently is.

 

[InstallGentoo]

issue with Miku NT

I differ, I really like NT. I didn't consider that it was a V2 sound alike. I consider it Miku but for me. V2-V4x was a bit before my time as a adult, (heck, v4x was when I got out of highschool).

 

[WacoWacko39]

I'm not certain if this the appropriate or best place to ask, but what is and "AI" voicebank? How is it any different from a more standard one? I am under the impression SynthVs are aforementioned. I'm fond of SynthV, though I'm only really familiar with Tetos 'sound'. And a fan I am, I can't help but feel she often sounds 'samey'; I would be remised, if for vertical movement they sacrificed lateral movement

 

[Bookworm2]

I'm pretty sure for SynthV at least that an AI learning program takes recordings of a voice provider singing and gets phonemes based on that algorithm. Not sure of the details though.

 

[IO+]

I hope her voice character doesn’t change much. The SP version has me a bit concerned.
Piapro NT already has inherent issues like abrupt silence, choppiness, and a tendency to randomly produce a thick vocal in an unpleasant way. This issues has been around since V2, though she’s improved a lot in terms of sound quality and overall usability—especially in V4, where she sounds really good.

Honestly, I find Miku V4 to be the peak of voice quality—what Miku should sound like.

 

[lIlI]

I'm not certain if this the appropriate or best place to ask, but what is and "AI" voicebank? How is it any different from a more standard one? I am under the impression SynthVs are aforementioned. I'm fond of SynthV, though I'm only really familiar with Tetos 'sound'. And a fan I am, I can't help but feel she often sounds 'samey'; I would be remised, if for vertical movement they sacrificed lateral movement

I'll do my best to explain!

As context for anyone who doesn't know, traditional vocal synths are made using the 'concatenative' method. A singer is hired to sing all the individual letter sounds of a language, called phonemes, at various pitches. These short recordings of vowels and consonants are then downloaded onto your computer when you buy a concatenative Vocaloid. Then, the Vocaloid software puts the phoneme recordings together to make the words you type into the program. It is a simple method that allows you to make a program that can say any word without you needing to record every word in the dictionary.

AI vocal synths use machine learning to create the voicebank. Instead of recording phonemes, the singer sings whole songs. Then you label where each phoneme is in each song, so the software knows how to read them. After labelling, the AI software learns the sonic qualities that give the singer their unique timbre and pronunciation. These rules are saved into a much smaller file, then applied dynamically by the software, making the output sound like the singer. Rather than simply playing a pre-recorded sound, the software calculates how a word should be sung by applying the rules it has learnt from 'listening' to the singer. Rather like a robotic impressionist. There's a lot more complex stuff happening under the hood, for example, you can feed the AI data from multiple different singers to increase the range of one voice, and make them sing fluently in multiple languages by mixing the 'rules' for English with the 'rules' for a non-native singer's timbre.

However, AI does not sacrifice your ability to make a voice sound unique. In fact, I believe AI voices are easier to edit and can achieve greater variety in performance compared to concatenative voices, because you're not restricted to the singer's recordings. If you push their settings, you can make one AI voice sound like multiple completely different people! The reason AI usage online can be samey is a by-product of convenience. AI voices sound natural out of the box, so producers feel less of a need to edit the results. Consequently, a lot of producers use AI voices at their default settings.

Most concatenative voicebanks sound a lot less polished by default, and require more hard work to sound natural. When you tune a concatenative voice, you can hear a big improvement in its results, which motivates producers to edit them by hand. This forces more users to develop their own unique style.