Talk:Lily/@comment-35874478-20181024182219/@comment-53539-20181027132614

Only differences is traits that apply to head and chest voices that happen to be recorded, they get transferred into Vocaloid and can be the deciding factor. Some singers head voices can sound like another singers chest voice in Vocaloid. It really doesn't matter either way as Vocalid just treats them all the same. As I said, only captured traits seem to matter. This is why we don't tend to worry about them as a classification. Len is a chest voice and Rin is a head voice, but Len can sound like Sweet Ann and as far as I know, she isn't a chest voice. So there is it in action pretty much. Either way, once transferred to Vocaloid even the Falsetto vocals can at times not matter. Its a problem with vocaloid being a synth. As a head and chest, different muscles should be considered and Vocaloid should have separate calculations, but it does not, so the only way to separate the vocals is pretty much lost and it treats them all the same.

Its pretty much as I said, why we have problems with falsetto vocals at times in Vocaloid. As far as Vocaloid is concerned, Falsetto is no different to chest and head vocals, or anything else. So this is why its all clumped together in such large groups "lower mature female", "mid mature female", "high mature female", "young female high", etc. Its also why there is a need to capture as many vocal traits into each voicebank, there is so little separating vocals those traits are the only thing that causes each Vocaloid to be different. Honestly after 8 years of learning about synths, most are the same. Chipspeech and Alter/ego are slightly different, their engines can support different synth processes (DeeKlatt and Dandy 704 don't make their vocals in the same way, one is live feedback and the other samples like Vocaloid), but most aren't *that* flexible. :-?

Eh, the nasally problem is well known so don't worry about it. I just explained how it happens, as what little knowledge exists on it. Joke or not, its a problem that can be quite annoying for Vocaloid. I await the day they finally rid it from the engine. And then all synths... I miss the days of COMPLETE BRO-KEN SYNTH FOR THIS REA-SON, as at least they were big, bold and often lacked the nasally sounds like the Atari's chipe Rotten.ST in Chipspeech is based on. Least you know its an established problem and the two pinned down suspected causes of it. I add my own take to the theory that is can be also half based on microphones, as many studios have this habit of relying on older mics. Most studios apparently use a microphone from the 80s I believe it was, forget which model. I once decided to go to the hassle of identifying the mics on boxarts such as Big Al's original Elvis-Frankenstein boxart and Meiko's. I found out from a article about it. CFM has, however, moved onto digital mics these days, and Nana was recorded with modern mics.

It was mostly VO forum'ers and odd fans both in Japan and West. IT was almost a post release comment as the original demos were pretty bad. When people finally got Lily they weren't expecting a deeper tone Vocal for this reason. Part of the blame is fans got used to Miku and a lot of V2 vocals keep being used like her. So when presented with a lower toned Lily, they failed to understand what they had. I doubt they have a optimum range given to them as well. IT was a growing concern about the demos being released because by the time Mayu English demo came out I did start seeing people question how useful demos were. Mayu was one of the worst, I myself could barely hear "English" and it was one of its biggest complaints. Another was Megpoid English did not have "fly me to the moon" yet Japanese had it in "English". But it started basically with Lily V2. Plus after Lily there was the start of the hissy-fit era wherein fans jumped on you if you dared compare Vocaloid vocals. Especially when VY1 came along and it got silly, "Miku Solid with a pinch of Rin" and that nonsense.

But yes, Lily did receive some Meiko comparisons, it was about 2010 it all started. I do often note at one stage I found myself at VO forums telling people about impersonators, as the knee-jerk reaction got really silly. There have been experiments done with two Vocaloids, one adjusted to sound more like the other. In cases like Rin and Len, or Mayu and IA, unless you know the basic traits it can happen that you end up mistaking them for each other. Its mostly down to tone so mentally you have to filter that out and look for ways like the way Vocaloids say things. Unless its obvious, like a language jump one, often even I struggle at times to tell certain Vocaloids apart in duets if their similar types. I don't think its bad either, honestly the human voice is based on a limit model and overlaps will happen.

Its not wrong to mistake one Vocaloid for another or think one sounds alike. It happens all the time with real singers and no fan should feel guilty or lack of pride if they fail to tell the difference. Embarrassing I suppose, but the Vocaloids don't care as their not real people, so its a saving grace.