VOCALOID Hatsune Miku VOCALOID6
- Thread starter MagicalMiku
- Start date
Miku Vocaloid6 will be available to freely try at Crypton booth of Niconico Chokaigi 2026 (April 25-26), together with Miku NT2, limited goods and..maybe some announcements too
(the hints are a bit vague, so the announcements might be about new goods, or some new figures, but might be also new features of Miku V6, NT2, or even some other Piapro Characters. We know that now the development team can focus more on Luka, but I think some announcements for her might happen during talking session of Magical Mirai in Tokyo, but it really depends on the development schedule
)
(the hints are a bit vague, so the announcements might be about new goods, or some new figures, but might be also new features of Miku V6, NT2, or even some other Piapro Characters. We know that now the development team can focus more on Luka, but I think some announcements for her might happen during talking session of Magical Mirai in Tokyo, but it really depends on the development schedule
)
I see most people on here are seemingly mostly positive for V6 Miku, so i thought I'd throw my voice into the ring for a more critical take on her vb.
I am not the biggest Miku V6 fan. She still sounds like NT, and i feel like Crypton's continuous chasing of that sound over something clearer is somewhat limiting them. I don't want another robotic Miku with clicky consonants as we already have SP and NT2. I would much rather have the AI voicebank actually sound like her most realistic option. V4X still reigns as her most used and bought VB even after NT and SP releases, and I think thats in part due to her more realistic tone, and while I know Crypton doesnt like realism, I have always believed that Miku's strength is being for everyone to use and have their own unique style with, and continuing to pursue the robotic NT sound is isolating those that would rather have a clearer, more V4X-like sound and more realistic style. While I don't exactly hate V6, I believe that the community giving too positive of a feedback to this will just continue to have Crypton remain in the NT sound and leave fans like myself, who prefer realistic vocals or hate NT, behind.
I am not the biggest Miku V6 fan. She still sounds like NT, and i feel like Crypton's continuous chasing of that sound over something clearer is somewhat limiting them. I don't want another robotic Miku with clicky consonants as we already have SP and NT2. I would much rather have the AI voicebank actually sound like her most realistic option. V4X still reigns as her most used and bought VB even after NT and SP releases, and I think thats in part due to her more realistic tone, and while I know Crypton doesnt like realism, I have always believed that Miku's strength is being for everyone to use and have their own unique style with, and continuing to pursue the robotic NT sound is isolating those that would rather have a clearer, more V4X-like sound and more realistic style. While I don't exactly hate V6, I believe that the community giving too positive of a feedback to this will just continue to have Crypton remain in the NT sound and leave fans like myself, who prefer realistic vocals or hate NT, behind.
Seconded. I listened to a bit of the english demo on this thread and she still sounded a bit too robotic for a V6. Compare that to the realism of V6 Gumi.
Not to say that I like SV approach of realism (I still do not like it that much), but at least CFM should learn about the mechanism behind the SV-like vocal realism, so that it can benefit CFM to produce a more individual character-tailored realistic voicebank, instead of modular realistic AI voicebank produced by SV
Crypton doesn't make the Vocaloid engine.
But CFM is still actively developing the Piapro NT studio am I right? They should put their efforts in studying all latest voice synth technologies without discriminating (in a prejudice way) how realistic or too "human-like" the vocal output is.
the thing is, crypton just doesn't seem to want to... i don't think they're avoiding more natural synthesis because it's too difficult or scary, or budgetary reasons, but rather that they've built up this big empire where miku and co are very specifically 'robotic virtual singers' - they're catering to the crowd of people who are probably more likely to say 'what's even the point of listening to sv covers/songs, i'd rather just listen to a real human atp'.
there are lots of people who clearly wouldn't care if the virtual singer-first approach was abandoned just for making a good modern voicebank, but they are very deadset on using different engines and synthesis styles to almost... preserve a charicature of what miku v2 was to people back in the day. if that makes sense in a weird roundabout way
they want to take the long route with NT so that they can make something stylistic and different, and it feels like yamaha's more recent push to make v6 also be an engine for that thing is probably what attracted them to v6. we don't know how, but crypton still helped out on parts of teto sv's development, so even if working backwards from the best and most natural result and shaping it into their synthetic vision would be better, again it feels like they're just (kinda stubbornly) dedicated to their very specific goal.
there are lots of people who clearly wouldn't care if the virtual singer-first approach was abandoned just for making a good modern voicebank, but they are very deadset on using different engines and synthesis styles to almost... preserve a charicature of what miku v2 was to people back in the day. if that makes sense in a weird roundabout way
they want to take the long route with NT so that they can make something stylistic and different, and it feels like yamaha's more recent push to make v6 also be an engine for that thing is probably what attracted them to v6. we don't know how, but crypton still helped out on parts of teto sv's development, so even if working backwards from the best and most natural result and shaping it into their synthetic vision would be better, again it feels like they're just (kinda stubbornly) dedicated to their very specific goal.
A special interview between Hatsune Miku V6 developer Wataru Sasaki and vocal tuner Bibi: The secret to rich vocal expression lies in "love for the character that sings" and "active fan activities"!? [PART1]
(translated from the original interview in japanese, available here)
https://plugplus.rittor-music.co.jp/learning/interview/miku-v6_sasaki-bibi/
The long-awaited Hatsune Miku V6 has finally been released! To celebrate, a special conversation has been arranged between development producer Wataru Sasaki and popular vocal tuner Bibi, whom he eagerly sought to talk to. We present a special conversation that is a must-see for Vocaloid fans, where they look back on the history of Vocaloid from their respective perspectives and passionately discuss the evolution leading up to V6 and their outlook for the future.
table of contents:
A special interview between Hatsune Miku V6 developer Wataru Sasaki and vocal tuner Bibi.
1)"This isn't a human!" I was really shocked.
2)"I'm tuning the voices with the desire to see the moment when the character shines."
3)V6 gives the impression of being an evolution from the "Miku-chan" feel I got from V2.
4)Hatsune Miku V6 Product Overview
A special interview between Hatsune Miku V6 developer Wataru Sasaki and vocal tuner Bibi.
■ Wataru Sasaki Profile:
Voice Team Manager, Crypton Future Media, Inc. Chief Producer of Hatsune Miku and related vocal synthesis projects. Born in Sapporo in 1979. Joined Crypton Future Media in 2005, and in 2007, was in charge of planning and developing the vocal synthesis software "Hatsune Miku," which became a huge hit. Since then, he has been involved in the production and direction of the company's vocal synthesis-related products and projects. In addition, he participates in the OngaACCEL project, which researches new music technologies, and has written liner notes for electronic music artists such as Aphex Twin, Arca, and Squarepusher.
■ Bibi Profile
A creator specializing in editing (tuning) vocals for vocal synthesis software. Since starting her activities in 2012, she has been in charge of vocal editing for songs produced by many Vocaloid producers such as Jin, *Luna, and Marasy. In addition, she has worked on the tuning of several virtual singer songs implemented in the smartphone game "Project Sekai Colorful Stage! feat. Hatsune Miku".
1)"This isn't a human!" I was really shocked.
Sasaki: When did you first become aware of Hatsune Miku, and what was the reason for that?
Bibi: I was originally interested in manga and anime, which were considered "subculture," and it was through that connection that a friend introduced me to Hatsune Miku, saying, "This is what's popular right now!" The first song I heard was "Miku Miku ni Shite Ageru♪" I remember being really shocked, thinking, "This isn't a human!" Since then, I've listened to various Vocaloid songs, seen singers and people who enjoy creating illustrations and videos based on the songs, and I've become very drawn to the culture centered around Vocaloid.
Sasaki: It's true that this culture originated on Nico Nico Douga, but back then, there was also the background of internet culture like 2chan. I remember a sense of celebration, like, "Wow, something this interesting has come out on the internet!" We even waved leeks as a symbol of joy. So, Bibi-san, you went through that phase too.
Bibi: That's right. Until then, the internet was simply something I consumed, but since I had originally been drawing illustrations, it stimulated my desire to "create something myself!"
Sasaki: It was a time when everyone was listening to Hatsune Miku because they wanted to hear her. Then came the "Melt Shock," and we got a sense of family with Kagamine Rin & Len, Megurine Luka could sing in English, Megpoid was a bit of a rival, and oh yeah, Gakupoid was a shock too... Back then, there was a thrilling sense of not knowing what would happen and how everyone would use the characters. When Megpoid came out, I was surprised by the charm of her singing voice, which was completely different from Hatsune Miku's. Bibi, who are your favorite virtual singers and creators?
Bibi: My absolute favorite is Kagamine Len (laughs). That's one of the reasons why I love Giga. Also, cosMo@BousouP is a pioneer in making Vocaloids sing in a unique way. The impact I felt when I first heard "The Disappearance of Hatsune Miku" was incredible. I had never heard a song like that before, so I felt the immense potential of Vocaloid. I still love it to this day.
Sasaki: I see. That kind of edgy impact is unforgettable, isn't it?
2)"I'm tuning the voices with the desire to see the moment when the character shines."
Sasaki: When I listen to the vocals that Bibi has worked on, I get the impression that the intention is clear and easy to understand, and that the person who did the tuning is someone with a deep understanding of vocal and musical expression. Earlier we talked a little about "singers", but how did you become a "vocal tuning specialist"?
Bibi: I've always loved music and singing, but more than anything, I love characters. Ever since I learned about the expressive technique of covering songs with Vocaloid, my desire to express my favorite songs with these characters has grown even stronger. Now, I listen to human songs and think every day about how I could express them using the VOCALOID Editor. I've studied the project files that KannazukiP has made publicly available so many times that I practically wore them out.
Sasaki: I see. Bibi's love for the characters, or rather, her desire to see them come alive, is what connects to the charm of her singing. I can even sense her determination to realize the expression she loves.
Bibi: I think my desire to see the moment when that character shines is what ultimately led me to specialize in voice tuning. If I liked music, I could pursue other paths like composing, but I love voices, and I want to pursue how that character can shine the brightest. If I had to put it into words, it would be "fandom activities". I do voice tuning as a form of fandom activity, so I think I'll continue doing it even if I end up alone in the future (laughs).
Sasaki: Perhaps that's what makes Bibi so appealing to listeners. The characters Bibi likes are also characters that everyone likes, aren't they? It feels like the nuances that everyone loves and wants to hear resonate with Bibi's feelings and become embodied in her music. For example, in Marasy's "Mugen no Ticket," there's a moment where you feel like, "This is where KAITO's eyes suddenly opened wide". Of course, Marasy's song is wonderful, but you can also feel the dynamism of the character's energy coming alive through his voice.
Bibi: Thank you. However, my style of work is based on the fundamental premise that I borrow everything from other people. That's why I want to cherish the songs and characters. In the end, that might lead to what everyone is looking for.
Sasaki: Bibi's vocal tuning, while using Vocaloid, also has a very lifelike quality. Do you ever expand on your vocal tuning ideas by singing yourself?
Bibi: Although I only sing as a hobby, when I first started, I would adjust the vocals while singing. I would try singing along, thinking about how to reproduce the original, and it felt like I was singing together with the song.
Sasaki: This might be the first time I've heard someone describe the feeling of "singing together." Singing along with the characters...that sounds great. Bibi, you really must love the characters.
Bibi: Yes, that's all it took to get this far (laughs).
Sasaki: Conversely, are there any specific points you focus on when tuning the voice to bring out the "Vocaloid-like" quality?
Bibi: I think it's about not drastically altering the dynamics. With humans, there's quite a bit of variation in vocal volume, but because it's not human, that difference isn't as pronounced. I think that's what makes it sound like a Vocaloid.
Sasaki: Bibi-san, it seems you also like that "saturated voice" feeling that's characteristic of Vocaloid, or that "overload" feeling where the Vocaloid's voice gets louder due to the volume difference at the transitions.
Bibi: I really love it. That's why I always set my attack power to a high level. My absolute favorite is Kagamine Rin & Len's Act 1 (the nickname for the first released version). That's the image I have of Vocaloid.
Sasaki: For Kagamine Rin & Len act1, we recorded their voices dynamically, partly to differentiate them from Hatsune Miku. We asked Asami Shimoda to push her voice to its absolute limits, and that resulted in that overloaded feeling. It wasn't an easy style to control, but it certainly had that dynamic quality that's characteristic of Vocaloid.
(translated from the original interview in japanese, available here)
https://plugplus.rittor-music.co.jp/learning/interview/miku-v6_sasaki-bibi/
The long-awaited Hatsune Miku V6 has finally been released! To celebrate, a special conversation has been arranged between development producer Wataru Sasaki and popular vocal tuner Bibi, whom he eagerly sought to talk to. We present a special conversation that is a must-see for Vocaloid fans, where they look back on the history of Vocaloid from their respective perspectives and passionately discuss the evolution leading up to V6 and their outlook for the future.
table of contents:
A special interview between Hatsune Miku V6 developer Wataru Sasaki and vocal tuner Bibi.
1)"This isn't a human!" I was really shocked.
2)"I'm tuning the voices with the desire to see the moment when the character shines."
3)V6 gives the impression of being an evolution from the "Miku-chan" feel I got from V2.
4)Hatsune Miku V6 Product Overview
A special interview between Hatsune Miku V6 developer Wataru Sasaki and vocal tuner Bibi.
■ Wataru Sasaki Profile:
Voice Team Manager, Crypton Future Media, Inc. Chief Producer of Hatsune Miku and related vocal synthesis projects. Born in Sapporo in 1979. Joined Crypton Future Media in 2005, and in 2007, was in charge of planning and developing the vocal synthesis software "Hatsune Miku," which became a huge hit. Since then, he has been involved in the production and direction of the company's vocal synthesis-related products and projects. In addition, he participates in the OngaACCEL project, which researches new music technologies, and has written liner notes for electronic music artists such as Aphex Twin, Arca, and Squarepusher.
■ Bibi Profile
A creator specializing in editing (tuning) vocals for vocal synthesis software. Since starting her activities in 2012, she has been in charge of vocal editing for songs produced by many Vocaloid producers such as Jin, *Luna, and Marasy. In addition, she has worked on the tuning of several virtual singer songs implemented in the smartphone game "Project Sekai Colorful Stage! feat. Hatsune Miku".
1)"This isn't a human!" I was really shocked.
Sasaki: When did you first become aware of Hatsune Miku, and what was the reason for that?
Bibi: I was originally interested in manga and anime, which were considered "subculture," and it was through that connection that a friend introduced me to Hatsune Miku, saying, "This is what's popular right now!" The first song I heard was "Miku Miku ni Shite Ageru♪" I remember being really shocked, thinking, "This isn't a human!" Since then, I've listened to various Vocaloid songs, seen singers and people who enjoy creating illustrations and videos based on the songs, and I've become very drawn to the culture centered around Vocaloid.
Sasaki: It's true that this culture originated on Nico Nico Douga, but back then, there was also the background of internet culture like 2chan. I remember a sense of celebration, like, "Wow, something this interesting has come out on the internet!" We even waved leeks as a symbol of joy. So, Bibi-san, you went through that phase too.
Bibi: That's right. Until then, the internet was simply something I consumed, but since I had originally been drawing illustrations, it stimulated my desire to "create something myself!"
Sasaki: It was a time when everyone was listening to Hatsune Miku because they wanted to hear her. Then came the "Melt Shock," and we got a sense of family with Kagamine Rin & Len, Megurine Luka could sing in English, Megpoid was a bit of a rival, and oh yeah, Gakupoid was a shock too... Back then, there was a thrilling sense of not knowing what would happen and how everyone would use the characters. When Megpoid came out, I was surprised by the charm of her singing voice, which was completely different from Hatsune Miku's. Bibi, who are your favorite virtual singers and creators?
Bibi: My absolute favorite is Kagamine Len (laughs). That's one of the reasons why I love Giga. Also, cosMo@BousouP is a pioneer in making Vocaloids sing in a unique way. The impact I felt when I first heard "The Disappearance of Hatsune Miku" was incredible. I had never heard a song like that before, so I felt the immense potential of Vocaloid. I still love it to this day.
Sasaki: I see. That kind of edgy impact is unforgettable, isn't it?
2)"I'm tuning the voices with the desire to see the moment when the character shines."
Sasaki: When I listen to the vocals that Bibi has worked on, I get the impression that the intention is clear and easy to understand, and that the person who did the tuning is someone with a deep understanding of vocal and musical expression. Earlier we talked a little about "singers", but how did you become a "vocal tuning specialist"?
Bibi: I've always loved music and singing, but more than anything, I love characters. Ever since I learned about the expressive technique of covering songs with Vocaloid, my desire to express my favorite songs with these characters has grown even stronger. Now, I listen to human songs and think every day about how I could express them using the VOCALOID Editor. I've studied the project files that KannazukiP has made publicly available so many times that I practically wore them out.
Sasaki: I see. Bibi's love for the characters, or rather, her desire to see them come alive, is what connects to the charm of her singing. I can even sense her determination to realize the expression she loves.
Bibi: I think my desire to see the moment when that character shines is what ultimately led me to specialize in voice tuning. If I liked music, I could pursue other paths like composing, but I love voices, and I want to pursue how that character can shine the brightest. If I had to put it into words, it would be "fandom activities". I do voice tuning as a form of fandom activity, so I think I'll continue doing it even if I end up alone in the future (laughs).
Sasaki: Perhaps that's what makes Bibi so appealing to listeners. The characters Bibi likes are also characters that everyone likes, aren't they? It feels like the nuances that everyone loves and wants to hear resonate with Bibi's feelings and become embodied in her music. For example, in Marasy's "Mugen no Ticket," there's a moment where you feel like, "This is where KAITO's eyes suddenly opened wide". Of course, Marasy's song is wonderful, but you can also feel the dynamism of the character's energy coming alive through his voice.
Bibi: Thank you. However, my style of work is based on the fundamental premise that I borrow everything from other people. That's why I want to cherish the songs and characters. In the end, that might lead to what everyone is looking for.
Sasaki: Bibi's vocal tuning, while using Vocaloid, also has a very lifelike quality. Do you ever expand on your vocal tuning ideas by singing yourself?
Bibi: Although I only sing as a hobby, when I first started, I would adjust the vocals while singing. I would try singing along, thinking about how to reproduce the original, and it felt like I was singing together with the song.
Sasaki: This might be the first time I've heard someone describe the feeling of "singing together." Singing along with the characters...that sounds great. Bibi, you really must love the characters.
Bibi: Yes, that's all it took to get this far (laughs).
Sasaki: Conversely, are there any specific points you focus on when tuning the voice to bring out the "Vocaloid-like" quality?
Bibi: I think it's about not drastically altering the dynamics. With humans, there's quite a bit of variation in vocal volume, but because it's not human, that difference isn't as pronounced. I think that's what makes it sound like a Vocaloid.
Sasaki: Bibi-san, it seems you also like that "saturated voice" feeling that's characteristic of Vocaloid, or that "overload" feeling where the Vocaloid's voice gets louder due to the volume difference at the transitions.
Bibi: I really love it. That's why I always set my attack power to a high level. My absolute favorite is Kagamine Rin & Len's Act 1 (the nickname for the first released version). That's the image I have of Vocaloid.
Sasaki: For Kagamine Rin & Len act1, we recorded their voices dynamically, partly to differentiate them from Hatsune Miku. We asked Asami Shimoda to push her voice to its absolute limits, and that resulted in that overloaded feeling. It wasn't an easy style to control, but it certainly had that dynamic quality that's characteristic of Vocaloid.
A special interview between Hatsune Miku V6 developer Wataru Sasaki and vocal tuner Bibi: The secret to rich vocal expression lies in "love for the character that sings" and "active fan activities"!? [PART2]
(translated from the original interview in japanese, available here)
https://plugplus.rittor-music.co.jp/learning/interview/miku-v6_sasaki-bibi/
3)V6 gives the impression of being an evolution from the "Miku-chan" feel I got from V2.
Sasaki: Bibi-san, you clearly remember the V2 version of Hatsune Miku, and now you've tried V6. Do you notice any differences between each version from V2 to V6?
Bibi: I feel that the differences are quite clear. V2 felt like "Miku-chan", but with V4X, she became "Miku-san", and I got the impression that she was becoming more and more like an older sister. However, there's always that reassuring feeling that "Hatsune Miku" will always be "Hatsune Miku". While synthesized voices are constantly evolving and becoming closer to human voices, I really feel that they are cherishing the original voice.
Sasaki: Did you feel that V6 was an extension of "Miku" this time around?
Bibi: The vocal tones of V6 and NT give me the impression that they are more of an evolution from V2's "Miku-chan." Rather than trying to make the voice sound more human, I think they carefully preserved and evolved Hatsune Miku's original vocal quality. I'd like to try re-recording songs I've already tuned using V6 to see what happens. I feel that the differences I observe will help me establish the direction of my own expression.
Sasaki: "Hatsune Miku V6" was developed with the image of Hatsune Miku as the virtual singer that everyone envisions. We prepared "original" and "soft" voice banks, and for the original, we mixed in a small amount of the "LIGHT" component from the V2 expansion voice library pack "Hatsune Miku Append" by breaking it down, and for the soft voice, we added a little of the "Sweet" component. We aimed to balance the quirks of VOCALOID6 with the brightness and calmness of the voice to create the Hatsune Miku we envisioned. At the time of development, LIGHT was a bit difficult to differentiate from the original V2 voice bank.
Bibi: I loved "LIGHT" and "VIVID" from "Hatsune Miku Append". I'm so happy!
Sasaki: At the time, I challenged myself to develop "Hatsune Miku Append" while thinking, "What would another 'Hatsune Miku' be like?" There was a reputation that "the original is the best after all," but it's deeply moving for me to see that it lives on in V6 over time. The image of Miku's voice as a character may be something that can be pursued endlessly. What are the differences between the original and the software?
Bibi: The original version had a lively feel and was easy to express. Previous versions, when left as is, had a somewhat subdued impression, but the original has added brightness, making it really fun to sing. The software is also easy to use and blends well with the song. Voices with a lot of breathiness tend to produce noise, but that wasn't the case here.
Sasaki: Because of the nature of Vocaloid, the voice can sometimes sound muffled or weak and get buried in the instrumental track, right? I've always felt bad about that for the creators, and I want to do something to make it more balanced and easier to hear. On the other hand, some people say that the almost fading voice was one of Vocaloid's charms. What are your thoughts on balancing ease of use with the individuality of the character?
Bibi: Certainly, previous Vocaloids had their imperfections, but from the perspective of someone tuning them, I always considered those imperfections to be part of their individuality. I think those imperfections were definitely part of their charm. Not just with Vocaloid, but with synthesized voices in general, I feel that the number of uncertain elements has decreased as technology has evolved. In the past, unexpected things would happen, like "If you write this phonetic symbol like this, it produces this sound", which was exciting (laughs). But with the evolution, the barrier to entry has definitely lowered, and I think it's become easier for beginners to use. However, it's a little sad that there's less room to play around with phonetic symbols as a result.
Sasaki: I see. It would be great if we could use AI to introduce a gimmick that deliberately produces outliers in vowels and expressions while maintaining clarity. It might be interesting if we could generate new pronunciations from random phonetic symbols, or forcibly combine consonants and vowels.
Bibi: Yes, that's right. Actually, I find "Megurine Luka V4X" really fun. It has so many different voice options, it's like a toy box.
Sasaki: That's how I felt at the time. Looking back, I think it might have been a bit excessive considering ease of use... (laughs).
Bibi: I kind of had a feeling that might be the case, but as a user, I was excited thinking, "With so many options, I'll be spoiled for choice forever!"
Sasaki: To hear you say that now makes me feel better (laughs). I also think it would be great if the world of songs sung by characters, including some quirky ones, could expand. I'm sure that one of the reasons Bibi-san does voice tuning is because she wants to hear all sorts of songs.
Bibi: Yes, I'd like to listen to a variety of songs. V6 has become very easy to understand and intuitive to use, so I hope that in the future, by utilizing AI and other technologies, it will be possible to create a wide variety of expressions.
4)Hatsune Miku V6 Product Overview
This is virtual singer compatible with VOCALOID6 software. It supports not only Japanese but also English, and Chinese support is planned for a free update in 2027 (MagicalMiku note: this is first time they confirm 2027 for the Chinese voicebank). Two versions are available: a starter pack that includes VOCALOID6 Editor (24,200 yen) and a voice bank for users who already own VOCALOID6 Editor (10,780 yen). Both versions include the DAW software Cubase LE. It is compatible with Mac/Windows and operates in AU/VST3 format.
Utilizing VOCALOID6 engine, the vocal synthesis has been optimized. By simply inputting lyrics and melody, natural breathing is automatically added, and fluent expression blending multiple languages is now possible. Furthermore, thanks to machine learning, it learns pitch changes, volume changes, timing, and accents, making it adept at creating vocal expressions that feel warm and naturally fluctuating, similar to a human voice. In addition, a new TAKE function has been added that generates multiple vocals with different nuances from a single phrase, making it easy to create harmony parts.
While the default settings now allow for natural-sounding singing, it's still possible to fine-tune parameters as before. The "Expression parameter" allows to adjust the intensity of the voice, and you can intuitively edit subtle vocal expressions such as scooping, vibrato, and pauses while viewing the pitch and volume curves on the screen.
(translated from the original interview in japanese, available here)
https://plugplus.rittor-music.co.jp/learning/interview/miku-v6_sasaki-bibi/
3)V6 gives the impression of being an evolution from the "Miku-chan" feel I got from V2.
Sasaki: Bibi-san, you clearly remember the V2 version of Hatsune Miku, and now you've tried V6. Do you notice any differences between each version from V2 to V6?
Bibi: I feel that the differences are quite clear. V2 felt like "Miku-chan", but with V4X, she became "Miku-san", and I got the impression that she was becoming more and more like an older sister. However, there's always that reassuring feeling that "Hatsune Miku" will always be "Hatsune Miku". While synthesized voices are constantly evolving and becoming closer to human voices, I really feel that they are cherishing the original voice.
Sasaki: Did you feel that V6 was an extension of "Miku" this time around?
Bibi: The vocal tones of V6 and NT give me the impression that they are more of an evolution from V2's "Miku-chan." Rather than trying to make the voice sound more human, I think they carefully preserved and evolved Hatsune Miku's original vocal quality. I'd like to try re-recording songs I've already tuned using V6 to see what happens. I feel that the differences I observe will help me establish the direction of my own expression.
Sasaki: "Hatsune Miku V6" was developed with the image of Hatsune Miku as the virtual singer that everyone envisions. We prepared "original" and "soft" voice banks, and for the original, we mixed in a small amount of the "LIGHT" component from the V2 expansion voice library pack "Hatsune Miku Append" by breaking it down, and for the soft voice, we added a little of the "Sweet" component. We aimed to balance the quirks of VOCALOID6 with the brightness and calmness of the voice to create the Hatsune Miku we envisioned. At the time of development, LIGHT was a bit difficult to differentiate from the original V2 voice bank.
Bibi: I loved "LIGHT" and "VIVID" from "Hatsune Miku Append". I'm so happy!
Sasaki: At the time, I challenged myself to develop "Hatsune Miku Append" while thinking, "What would another 'Hatsune Miku' be like?" There was a reputation that "the original is the best after all," but it's deeply moving for me to see that it lives on in V6 over time. The image of Miku's voice as a character may be something that can be pursued endlessly. What are the differences between the original and the software?
Bibi: The original version had a lively feel and was easy to express. Previous versions, when left as is, had a somewhat subdued impression, but the original has added brightness, making it really fun to sing. The software is also easy to use and blends well with the song. Voices with a lot of breathiness tend to produce noise, but that wasn't the case here.
Sasaki: Because of the nature of Vocaloid, the voice can sometimes sound muffled or weak and get buried in the instrumental track, right? I've always felt bad about that for the creators, and I want to do something to make it more balanced and easier to hear. On the other hand, some people say that the almost fading voice was one of Vocaloid's charms. What are your thoughts on balancing ease of use with the individuality of the character?
Bibi: Certainly, previous Vocaloids had their imperfections, but from the perspective of someone tuning them, I always considered those imperfections to be part of their individuality. I think those imperfections were definitely part of their charm. Not just with Vocaloid, but with synthesized voices in general, I feel that the number of uncertain elements has decreased as technology has evolved. In the past, unexpected things would happen, like "If you write this phonetic symbol like this, it produces this sound", which was exciting (laughs). But with the evolution, the barrier to entry has definitely lowered, and I think it's become easier for beginners to use. However, it's a little sad that there's less room to play around with phonetic symbols as a result.
Sasaki: I see. It would be great if we could use AI to introduce a gimmick that deliberately produces outliers in vowels and expressions while maintaining clarity. It might be interesting if we could generate new pronunciations from random phonetic symbols, or forcibly combine consonants and vowels.
Bibi: Yes, that's right. Actually, I find "Megurine Luka V4X" really fun. It has so many different voice options, it's like a toy box.
Sasaki: That's how I felt at the time. Looking back, I think it might have been a bit excessive considering ease of use... (laughs).
Bibi: I kind of had a feeling that might be the case, but as a user, I was excited thinking, "With so many options, I'll be spoiled for choice forever!"
Sasaki: To hear you say that now makes me feel better (laughs). I also think it would be great if the world of songs sung by characters, including some quirky ones, could expand. I'm sure that one of the reasons Bibi-san does voice tuning is because she wants to hear all sorts of songs.
Bibi: Yes, I'd like to listen to a variety of songs. V6 has become very easy to understand and intuitive to use, so I hope that in the future, by utilizing AI and other technologies, it will be possible to create a wide variety of expressions.
4)Hatsune Miku V6 Product Overview
This is virtual singer compatible with VOCALOID6 software. It supports not only Japanese but also English, and Chinese support is planned for a free update in 2027 (MagicalMiku note: this is first time they confirm 2027 for the Chinese voicebank). Two versions are available: a starter pack that includes VOCALOID6 Editor (24,200 yen) and a voice bank for users who already own VOCALOID6 Editor (10,780 yen). Both versions include the DAW software Cubase LE. It is compatible with Mac/Windows and operates in AU/VST3 format.
Utilizing VOCALOID6 engine, the vocal synthesis has been optimized. By simply inputting lyrics and melody, natural breathing is automatically added, and fluent expression blending multiple languages is now possible. Furthermore, thanks to machine learning, it learns pitch changes, volume changes, timing, and accents, making it adept at creating vocal expressions that feel warm and naturally fluctuating, similar to a human voice. In addition, a new TAKE function has been added that generates multiple vocals with different nuances from a single phrase, making it easy to create harmony parts.
While the default settings now allow for natural-sounding singing, it's still possible to fine-tune parameters as before. The "Expression parameter" allows to adjust the intensity of the voice, and you can intuitively edit subtle vocal expressions such as scooping, vibrato, and pauses while viewing the pitch and volume curves on the screen.