To speak from the heart here, and as someone who's observed the development of Vocaloid for about a decade now, and knows a little about vocalsynth dev from the perspective of an UTAU dev and has talked to people with experience in this, I have to say that I cannot emphasize enough how much people take for granted how limited Crypton has historically been under Yamaha, and how much Crypton no longer being under them raises the ceiling of possibility just out of the sheer principle of not being subject to a level of bureaucratic hell.
Example: The concept of "VB updates" as we currently know it is actually a byproduct of Yamaha's honestly rather draconian control over how VBs are released. One of the first things Crypton did after the announcement was
update Miku Chinese. That probably passed over non-Chinese users' heads like "oh, whatever", but that is actually a
really big deal. Why is that a big deal? Because
Yamaha doesn't let you patch Vocaloid libraries. That doesn't sound like that much normally, because we're used to seeing a VB get released in its final form on release day, and any problems it ships with are permanent...but in the grand scheme of things, that's...extremely abnormal. Hell, Yamaha will even patch its
own engine, but they won't let its licensees submit patches like, well, literally every other software on the planet should have the right to do. Other engine licensors are generally not nearly as draconian about that, either (I regularly get Kontakt library updates pushed with no problem).
Under normal circumstances, when you have a software, it's not expected that it's going to ship bug-free! That goes for any kind of digital software, even video games. You can only test behind the scenes so much, and once it ships out to userland, the fact it's going to be scrutinized under thousands of users will inevitably reveal problems that weren't caught in dev. And under said normal circumstances, once the devs hear about it, they can work on fixing the problems and ship an update after the fact. And if halfway down the line, the devs come up with an idea for a great new feature that people would certainly like, they can add it and ship it as a free update to all of the users without having to put out a whole new market product. This is the same for NT; in addition to the prototype period, even if Miku NT ships in August 2020 not being the best product in the world, Crypton can continue to evaluate user feedback and continue to refine it further as per the users' needs, and if wat comes up with another idea on the fly, they can add that in too, and
then maybe increment the version number and put it out as a completely new iteration when they feel the update is particularly significant. You know, like...any other software.
And we
know Crypton has actually been limited in this fashion before. Proof: Rin and Len act2. To this day, act1 is the
only of their Vocaloid releases Crypton refuses to continue selling -- they even keep the V1s, and the other V2s, but they've shown every indication of wanting to actually wipe out act1 from existence. Why? Because act2, being a very minor touchup to act1 and not a full rerecording, was obviously
intended to be a postrelease patch in this manner. They've been wanting to do this since 2007. But Yamaha clearly did not
let them release it as a patch, which resulted in Crypton having to
physically mail a disc with act2 for a limited time to any existing act1 user, and then quietly discontinue act1 from sale. For another example, take the concept of E.V.E.C. -- like it or not, the fact of the matter is that it's an experimental feature Crypton clearly
wanted to have included in all their products, yet weren't able to implement in Yamaha's editor, and thus it's limited only to Piapro Studio. And when wat suggested the concepts of Miku Clear (which, in retrospect, is pretty obviously the beginnings of the concepts that led to the development of NT), the first reaction anyone following had was "how on earth is he going to convince Yamaha to let him do that?!" And now they're finally updating Miku Chinese...the one that just so happens to be the only one on the standalone Piapro Studio editor.
Take another example: think about how "appends" or separate libraries being actual separate market releases has been a thing taken for granted, with XSY being so favored because it's the only thing that lets you have much crossover between them, only for it to get unceremoniously canned in V5. On a wider level, let's consider: language support. You do have fringe cases like MAIKA and SeeU having a few "extra phonemes" to bring them closer to English, but otherwise, such cases are generally very rare. Why? Because of Yamaha's enforcement -- you can't patch a VB to have extra phonemes if you feel maybe you want to add those in, and for the most part they enforce a very strict separation of "one VB per language". Let's say you wanted to add support for Romance languages to a Japanese Vocaloid; it should be an easy update, Japanese has the same five vowels anyway, so all you have to do is tack on a few extra consonants and you're good, right? Except that still sends you into Bureaucracy Hell; Yamaha won't let you patch the VB, so you have to actually push it as an entirely separate library, with all the licensing labor and fees that come with it, and because it'll be so similar to the original product, your sales might not be so high because your target audience is likely to go "well, I'll just stick with the original, why should I buy a whole new VB?"
(Meanwhile, when non-Japanese UTAU development was in its early stages in the late 2000s, one of the first things attempted was a concept of a VB that could approximate multiple languages at once. It didn't work out, largely due to limitations with UTAU and SHIFT-JIS encoding and enforcement of morae-style recording, but I still think it was a perfectly fine idea that could have been fantastic if conditions had been better.)
I can't necessarily claim that Crypton will necessarily take advantage of all of the possibilities affronted to them, and even if they do try, perhaps they won't succeed, and in the end nothing in this world is perfect -- more importantly, nothing in this world will please everyone. (I personally
vastly prefer the new tone of NT to the Vocaloid engine's, as someone who has had multiple gripes with how said engine has a nasty tendency to just make everything horribly flat and difficult to work with. But I'm well aware that it's a turnoff to many people, and that it's probably impossible to create a sound that will be loved by everyone universally.) But as someone who's witnessed what very much looks to be Crypton clearly
wanting to break out of their shell and wanting to do more experimental, interesting stuff and work more with the userbase, and being limited by what seems to be some very strict and horribly arbitrary limitations put on them for no reason other than Bureaucratic Hell, it's not hard to see why they would make this decision, and it's much easier to see how there might be potential benefits in the long run that aren't immediately obvious -- for both the devs and the actual users.
--
Some miscellaneous notes:
- In regards to the Internet requirement, "always" is a pretty horribly abbreviated and bad translation of what they're referring to. An Internet connection requirement of some sort has been listed on pretty much every one of Crypton's Vocaloid releases since V3, so it's probably referring specifically to the activation process. That "always", on the Japanese page, actually says "常時接続", a Japanese technical term that roughly means "don't use dial-up, please". There's precedent for that kind of sentiment, too; Rin/Len V4X for Mac apparently really wants you to have a broadband Internet connection.
- While Twindrill and Crypton are on good terms and an experiment to see Teto's performance on the new engine isn't out of the question, I think the bigger question of whether Crypton will want to take her in is, ironically, more related to bureaucracy. In the end, Twindrill most likely wants to retain full rights and responsibility for Teto, and Crypton most likely doesn't want to be saddled with the extra responsibility of taking care of her, along with both of them probably wanting to set clear boundaries that she's not actually a Piapro Character. So I feel it's unlikely she'll end up a full-on product on it, and even if they do, I expect they'll be taking significant measures to keep them separated in a branding sense. (As a fun point, though, in the Teto 10th Anniversary livestream where they discussed sending Teto's samples to Crypton, they confirmed it was just something they did For Science, but also said that it would be funny if she could somehow turn into some kind of Vocaloid/UTAU hybrid -- then she'd be a "real chimera".)
