Board Thread:Questions and Answers/@comment-89.144.193.214-20150305164351/@comment-25521131-20150525173803

107.161.199.32 wrote: @CursedMyth Crypton already said All their V3/V4's were new recordings, not to mention Miku original's pronounce and pitch is nothing like her V2, she's nearly two coctaves lower and has a more mature tone. The V2>V3 thing has already been debunked. (Not to mention Crypton said Miku V3 was freash recordings...Right when they anounced her. :/) And where exactly did they announce that?

All the stuff you mentioned could have easily been done with editing the results of the samples from the original vocal, something Crypton has been known to do. Have you ever noticed how straight Crypton's vocals are in comparison to most other Vocaloids? That's likely the result of internally editing the samples where they straighten out the notes of the actual samples of the vocal. Kasane Teto also incorperates this, intending to sound like a Crypton vocal, but in her sampling, her vocals have been edited and straightened, as one can clearly hear, the samples are not raw vocals, it's the same case with Crypton Vocaloids. Then we have things like Luka V4x ENGLISH SOFT, it's not definite, but it's difficult to believe that ENGLISH SOFT wasn't just an edit of ENGLISH STRAIGHT - especially when you consider that they have the exact same phonetic errors, such as [@r] [l] and [@U] being pronounced incorrectly, but between the two voicebanks are pronounced exactly the same, and it makes sense, as if they just edited one vocal from the other, they could use the exact same settings they used to tune the original STRAIGHT vocal, and they wouldn't have to work on yet another English vocal - and with that in mind, considering how easy it is to make a mistake with phonetics in English vocals, making 2 completely seperate English vocals, and allowing them to cross-synthesize with each other would be a difficult task, especially for a Japanese company, seeing as if there was a phonetic error or mispronunciation in one voicebank, but not the other, it wouldn't be able to cross-synthesize properly, so as horrible as the voicebank sounds, it actually makes a lot of sense as to how and why the ENGLISH SOFT voicebank would have just been an edit of the ENGLISH STRAIGHT voicebank, which vrings me back to my original point. Crypton has been known to do a significant amount of editing in the recorded voicebank samples - in fact, it's suggested they intentionally edited the consonants for V3 since they recently posted a sound on soundcloud comparing the consonants from Miku V3 and the edited version of the same consonant to be used in Miku V4 with the intention of Miku V4's consonants to sound more like her V2 version. So unless they have definitively said that Miku V3 is a re-recording, which if they have, please do correct me, then there is the possibility, that it's just an edit of the original V2 vocal. You also might want to consider, that whever they record an update to an original vocal, they've called it "STRAIGHT", examples being KAITO STRAIGHT, MEIKO STRAIGHT, and LUKA ENGLISH STRAIGHT, vocals we all definitively know have been re-recorded, and yet, Miku V3's main vocal is called Miku Original, which seems to imply it's the original vocal ( However, to be reasonable, regarding their naming system, they have also used the term "Native" before to describe Luka's V4 update in EVEC, so this could mean I'm just flat out wrong, but it's just odd that they didn't use STRAIGHT for Miku, but they used STRAIGHT for Luka )