And here I was holding out hope for Meiko/Kaito NT releases too. 
Guess it'll be a LONG time before they even think about the other 'loids at this point.
Guess it'll be a LONG time before they even think about the other 'loids at this point.Other Cryptonloid voicebank updates, collabs, & concert news (crypton_wat Twitter translations)
- Thread starter uncreepy
- Start date
Note: This post is REALLY messy because I am mega tired and I don't care to rewrite it into proper wording. orz
Miku NT update that was supposed to come last month is available:
The last time we got an update for NT Voice Library HATSUNE MIKU Original was December 25th, 2020 (1.0.0.4). Today's update mainly improved phoneme transitions in her voice and devoicing. F0 refers to the fundamental frequency related to the shape/expressiveness of the singing voice.
Last version of Piapro Studio NT was December 25th, 2020 (3.0.1.5). Today's update is 3.0.2.0, which corresponds to Miku NT Original 1.0.0.5.
I'm worried because for processing speed, it says "first stage", so I guess we can look forward to more processing speed updates because NT was REALLY SLOW. The update made the interface seem to run more smoothly so I am happier with that.
Weirdly, when you open NT and go to Help > Check for Updates..., it says it's up to date, but it's not.
To update Miku and the engine, go to your Sonicwire MY PAGE > LICENSED PRODUCT > scroll down to Windows/Mac Installer > Download the zip > Use Crypton Product Installer (both should have a checkmark on the install list, so just click Install on the pop-up).
I had problems with installing this. You allegedly don't have to uninstall either Miku or NT ahead of time because this does it for you, but mine got stuck (will explain). Engine installs 1st, then Miku. CryptonProductInstaller stayed stuck on "Finalizing Installation" for a ovver 20 minutes before I killed it with Task Manager. I highly recommend uninstalling Miku and NT BEFORE updating.
This is my thread documenting my problem with screenshots:
This is the solution:
You can finally drag tracks to rearrange them!
Miku NT update that was supposed to come last month is available:
The last time we got an update for NT Voice Library HATSUNE MIKU Original was December 25th, 2020 (1.0.0.4). Today's update mainly improved phoneme transitions in her voice and devoicing. F0 refers to the fundamental frequency related to the shape/expressiveness of the singing voice.
Last version of Piapro Studio NT was December 25th, 2020 (3.0.1.5). Today's update is 3.0.2.0, which corresponds to Miku NT Original 1.0.0.5.
I'm worried because for processing speed, it says "first stage", so I guess we can look forward to more processing speed updates because NT was REALLY SLOW. The update made the interface seem to run more smoothly so I am happier with that.
Weirdly, when you open NT and go to Help > Check for Updates..., it says it's up to date, but it's not.
To update Miku and the engine, go to your Sonicwire MY PAGE > LICENSED PRODUCT > scroll down to Windows/Mac Installer > Download the zip > Use Crypton Product Installer (both should have a checkmark on the install list, so just click Install on the pop-up).
I had problems with installing this. You allegedly don't have to uninstall either Miku or NT ahead of time because this does it for you, but mine got stuck (will explain). Engine installs 1st, then Miku. CryptonProductInstaller stayed stuck on "Finalizing Installation" for a ovver 20 minutes before I killed it with Task Manager. I highly recommend uninstalling Miku and NT BEFORE updating.
This is my thread documenting my problem with screenshots:
This is the solution:
You can finally drag tracks to rearrange them!
I messed around with Miku NT and documented (with screenshots) in this thread...
1) How to change E.V.E.C. colors
2) Noticed the Settings file includes Spanish, German, French, & Korean for Display Languages (I know Yukari had German but the others can't be used) and Default Lyrics included French
3) There's now a thing called Expression Style when you right click a track. It's Active by default, Normal is less curvy. I think Active sounds much more realistic/expressive.
I made Miku sound like a shower and it was disturbingly popular so I shared the NT file:
Made more Miku NT Engrish:
I know some people are unhappy with the newest NT update because they are crashing, but I haven't actually had much problems with it for once, so I'm happy with it running smoother. (I know I diss NT a lot but I do want it to end up being a good product because I paid money for it. orz)
1) How to change E.V.E.C. colors
2) Noticed the Settings file includes Spanish, German, French, & Korean for Display Languages (I know Yukari had German but the others can't be used) and Default Lyrics included French
3) There's now a thing called Expression Style when you right click a track. It's Active by default, Normal is less curvy. I think Active sounds much more realistic/expressive.
I made Miku sound like a shower and it was disturbingly popular so I shared the NT file:
Made more Miku NT Engrish:
I know some people are unhappy with the newest NT update because they are crashing, but I haven't actually had much problems with it for once, so I'm happy with it running smoother. (I know I diss NT a lot but I do want it to end up being a good product because I paid money for it. orz)
(Note: The machine translation on this wasn't entirely the best, as it's a little abstract. But I'll do my best to paraphrase accurately.)
Wat shared his impressions of how the virtual singers are shaping up, saying they currently fall into different fortes/general natures.
Miku
Good general purpose voice
Sounds thin
Rin and Len
Sound similar to how they do currently
Consonants are too strong
Luka
Soft voice
Can be weak sometimes
MEIKO
Her "a" and other vowel sounds make her voice "float" (Not sure what that means.)
Has a unique sound.
KAITO
Sharpness has improved
Somewhat choppy
These general descriptions are based on the work that has been done with them for various projects. Work related to Project Sekai...has led either to finding the particular listed weaknesses or refining the VBs to do better in particular areas, I can't tell which. Whichever it is, he mentions that it's related to Sekai and can be heard in different scales, pitch transitions, melody, speaking, tunes (tuning?), and other things.
Wat says there's still a long way to go, and he'd like to devote himself to that work.
If nothing else, assuming I'm right that these are NT-related, it seems to basically tell us that all the CFM 'loids are being developed in parallel, which was a question raised in this thread before.
Wat shared his impressions of how the virtual singers are shaping up, saying they currently fall into different fortes/general natures.
Miku
Good general purpose voice
Sounds thin
Rin and Len
Sound similar to how they do currently
Consonants are too strong
Luka
Soft voice
Can be weak sometimes
MEIKO
Her "a" and other vowel sounds make her voice "float" (Not sure what that means.)
Has a unique sound.
KAITO
Sharpness has improved
Somewhat choppy
These general descriptions are based on the work that has been done with them for various projects. Work related to Project Sekai...has led either to finding the particular listed weaknesses or refining the VBs to do better in particular areas, I can't tell which. Whichever it is, he mentions that it's related to Sekai and can be heard in different scales, pitch transitions, melody, speaking, tunes (tuning?), and other things.
Wat says there's still a long way to go, and he'd like to devote himself to that work.
If nothing else, assuming I'm right that these are NT-related, it seems to basically tell us that all the CFM 'loids are being developed in parallel, which was a question raised in this thread before.
I skimmed the tweets the other day, but it's NT related. I thought it implied they wanted to make them a little more versatile than their specialties compared to when they were Vocaloids. There are actually 2 more tweets he made yesterday.
One saying they would provide an update for flaws this week:
And one mentioning Project Sekai, Pocari Sweat, Yamaha, continuing to develop all Cryptonloids NTs:
As for the tweet machine translation can't handle:
The general classifications of Cryptonloids =
Miku: Though she is all purpose, she can also be used for fine/detailed tuning (that's where machine translation got "thin" from, 細い can mean small/minute, detail. 事 means things/matters).
Rin & Len: (Xさせ易い means "easy to X") Have voices that are raw and easy to accent, but there are times the consonants are too strong.
Luka: Soft voice quality, but has things she's not good at.
MEIKO: While her voice has uniqueness through things like her vowel traits (such as "a"), there are times her voice becomes unsteady (that's what 浮く"float" came from).
KAITO: His sharpness has improved, but there are times a feeling of choppiness remain.
(Reply kind of vague because still feel bad from irl stuff but just wanted to pop in and clear up about the voice qualities.)
One saying they would provide an update for flaws this week:
And one mentioning Project Sekai, Pocari Sweat, Yamaha, continuing to develop all Cryptonloids NTs:
As for the tweet machine translation can't handle:
The general classifications of Cryptonloids =
Miku: Though she is all purpose, she can also be used for fine/detailed tuning (that's where machine translation got "thin" from, 細い can mean small/minute, detail. 事 means things/matters).
Rin & Len: (Xさせ易い means "easy to X") Have voices that are raw and easy to accent, but there are times the consonants are too strong.
Luka: Soft voice quality, but has things she's not good at.
MEIKO: While her voice has uniqueness through things like her vowel traits (such as "a"), there are times her voice becomes unsteady (that's what 浮く"float" came from).
KAITO: His sharpness has improved, but there are times a feeling of choppiness remain.
(Reply kind of vague because still feel bad from irl stuff but just wanted to pop in and clear up about the voice qualities.)
This week's update for Piapro NT, mentioned 2 days ago, has arrived. It focuses on fixing editing issues, especially related to pitch curve editing. There's a more detailed description in the linked blog post.
This Tweet didn't really get translated well enough for me to feel comfortable relating it, but it sounds like they're continuing to work on the Dark/Whisper banks. Specifically, they're applying the technique for adjusting voice overtone, learned from MEIKO and KAITO, as well as a method for arranging the length of the voice frame. It's obviously been a wait (relative to the original release date, I assume), but Wat says things are progressing quickly at the moment. Some unspecified things are supposed to be finished by the end of May.
I just wish we could get some sort of demo / voice-preview... I think all we have got is like, that one second in Stella Sekai ver. for Dark, and even that isn’t 100% confirmed to be NT Dark+. With how Original+ “demos” sounded and the state they originally released the voicebank, I don’t understand why they don’t want to showcase neither of them.
Also, does Proseka have any example yet of what wat means with the “Voice Overtone” of KAITO and MEIKO?
Also, does Proseka have any example yet of what wat means with the “Voice Overtone” of KAITO and MEIKO?
I ranted about this on Twitter a while ago but forgot to write about it here, but Wat ALWAYS likes to name-drop a Cryptonloid periodically (especially to break of his long periods of silence) by saying things like "we're working on balancing Kaito this week, next week is Rin/Len" and used to love the word "soon". So to me, this tweet is absolutely useless because they never ever show audio or images to show anything significant/tangible.
"Harmonic tuning method" and "method for organizing voice length frames" is also fluffy wording that means nothing to most people, especially since work on it up until now has secretly been tested on Meiko/Kaito apparently and maybe not even revealed to the public. It just sounds like busywork. On top of that, he says their "goal is to complete various things during May", and everyone assumes it's Dark and/or Whisper because it was conveniently in the same sentence. With Crypton's extremely horrible track-record for never being on time and always having to push back deadlines once or twice, I seriously doubt we're going to get those voice banks next month. Even if we do, they're definitely not going to be completed, because Original is still getting constant updates. Also, "various things" could mean back-end interface stuff or bug updates like usual instead.
They're probably just bringing up an NT update because the last one came in March so they feel pressure to do another one "soon". But I would not be surprised at all if this "during May" update gets delayed like the rest of them. I would confidently bet money that Miku NT is still going to be getting updates when we hit her Christmas anniversary.
Edit: Oh yeah, they also said they were "hastily progressing" before apologizing for the wait, and I am sitting her thinking this entire project was rushed, we don't need more.
"Harmonic tuning method" and "method for organizing voice length frames" is also fluffy wording that means nothing to most people, especially since work on it up until now has secretly been tested on Meiko/Kaito apparently and maybe not even revealed to the public. It just sounds like busywork. On top of that, he says their "goal is to complete various things during May", and everyone assumes it's Dark and/or Whisper because it was conveniently in the same sentence. With Crypton's extremely horrible track-record for never being on time and always having to push back deadlines once or twice, I seriously doubt we're going to get those voice banks next month. Even if we do, they're definitely not going to be completed, because Original is still getting constant updates. Also, "various things" could mean back-end interface stuff or bug updates like usual instead.
They're probably just bringing up an NT update because the last one came in March so they feel pressure to do another one "soon". But I would not be surprised at all if this "during May" update gets delayed like the rest of them. I would confidently bet money that Miku NT is still going to be getting updates when we hit her Christmas anniversary.
Edit: Oh yeah, they also said they were "hastily progressing" before apologizing for the wait, and I am sitting her thinking this entire project was rushed, we don't need more.
I’m excited! Dark is probably the most acclaimed of Miku’s extra VBs, so I can’t wait to hear how it’ll sound in NT. I know they’re still trying to get their collective shit together, but given how many projects they’re undertaking on their own now, and given that 85% of their output prior has been 
Luka V4X and the first two Kagamines banks are outliers and should not be accounted for I trust things will turn out well.
Luka V4X and the first two Kagamines banks are outliers and should not be accounted for I trust things will turn out well.
This is... interesting. I'll stay dubiously hopeful for now! I'm interested in seeing how an append bank sounds in NT
Tbfh I still think Act 1 Kagamines are better quality than Act 2. Act 2 didn't fix like any pronunciation errors and honestly made them more muffled. Plus in V2 you can use the phoneme trick to smooth their vowel transitions anyway :P
This might be speculation but I think the job offer they put up in December was indeed for Proseka covers and other stuff, since the tuning and mixing of the “V Singers” suddenly got a lot better after around Ifuudoudou’s release (except Dreaming Chuchu wtf happened there), so I think wat and the development team now have more time to work on the voicebanks without being 100% dependent on Colorful Palette’s requests.I’m excited! Dark is probably the most acclaimed of Miku’s extra VBs, so I can’t wait to hear how it’ll sound in NT. I know they’re still trying to get their collective shit together, but given how many projects they’re undertaking on their own now, and given that 85% of their output prior has been
As someone who loved V2 era Miku Dark, I really hope this comes out well.
lol, look, it's just like I said. This week is Luka enhancing
(I assume he wrote 完ん全に by accident and meant 完全に)
I can’t wait to hear the changes to Luka’s voicebank made **this week** , **two months** from now in a 10 sec segment of a Project Sekai song.
Sometimes I wonder if it is about as specific as they can get. They show up for work with plans to enhance Luka. Change some things that don't sound quite right. Then, after ten hours of messing with audio equipment, editing samples, tweaking parameters, and adjusting algorithms they have nothing to show for it. They have just figured out a few hundred things that don't work and hope the next day will yield progress.
So the full version of ECHO came out today-ish and many are saying they used some sort of Luka NT in the song??? It’s either that or a REALLY weirdly tuned V4, but she also shows some differences in pronunciation / engine noise so idk.
I personally think Luka’s alpha or whatever bank has English capability by having extra phonemes ala
/ 
to avoid unnecessary work (she pronounces some words in a kinda funny way (like “black”) in the song so maybe that’s why).
(idk if this should go here, but given that Wat refuses to share proper demos/previews this may be our only actual clue about Luka’s development in months)
I personally think Luka’s alpha or whatever bank has English capability by having extra phonemes ala
/ to avoid unnecessary work (she pronounces some words in a kinda funny way (like “black”) in the song so maybe that’s why).(idk if this should go here, but given that Wat refuses to share proper demos/previews this may be our only actual clue about Luka’s development in months)
