Talk:Megurine Luka V4X/@comment-174.254.137.61-20180817014251/@comment-53539-20180817103547

They aren't the only studio that cuts corners. We're heard of Internet co doing the same thing. LUMi also mentioned that she is one of vocaloids not re-using samples from other sources and all samples are for the Vocaloid itself. Basically studios do things like hire one person to record a missing sound (they can hide it via sound engineering software), cut it from other sources by the same provider, etc. ITs all done because not all providers can be in the studio at the drop of a hat.

Whether or not you like it, Luka was indeed a test. Though Luka was in production since at least 2010, bare in mind it wasn't on V4x. First they were working on Appends, but Yamaha screwed them over and released V3. This caused all CFM products to have to be redone, hence CFM's late introduction into V3. They had nothing prepared. Internet co also got impacted, but were not as far into Megpoid Extend and were able to flip the project easily and were ready in time for the release. CFM had also a move to deal with as they went to a bigger office building. They also had Kaito V2, Meiko V2, Miku V2 English on the table in addition to Luka Append. All of this had to be redone. So Luka Append V2 was scrapped again. Kaito and Meiko went to the front and Miku ahead of Kagamine V3 and Luka V3.

Once again as they restart the Luka V3 project, Yamaha told the studios early V4 was on its way (they know half a year before we do). The new feature XSY, and this lead to EVEC. Unfortunately this also lead to Luka being only worked upon for a matter of months. They also didn't want to repeat history and Luka was rushed out in early V4, as CFM lost ground to other studios by their late entry into V3. The result is why Luka V4x is a bit messy. ITs a rush product with lots of ideas tossed in by not time to finalise them. Luka was the prototype or "test" vocal. You could blame Yamaha, but bare in mind... Yamaha is the bigger company here and is much bigger then CFM, with markets in transport and music among other such things. Whereas CFM is just a sound sample library company producer and distributor in Japan for such like companies such as the likes of Zero-G.

As far as we know, the only test releases for Miku were the Appends and Chinese voicebanks. The Appends were CFM's way of trying to repeat history and get another 60,000+ sales while keeping the fans hooked on Miku to sell Miku more. They were based on feed back from Act1 and Act2 Kagamines and the fact the differences between the two resulted in one set doing different songs to the other. The two Kagamines releases, however, were not tests. Chinese Miku is a test to see if more Vocaloids can be released into Chinese. Gumi had the same thing, her English was a test for Gackpo and Lily English (no show, no go?), so CFM again aren't the only company to do that. Generally, you soon learn a lot of things about a company over time. CFM for example would like it Sweet Ann wasn't released just before Miku, so they often neglect her when introducing Miku as the first V2 (they've had to re-word this carefully over time as a result to "First members of the CV series", "first Japanese Vocaloid", "First Vocaloid2 vocal for Vocaloid2" which is true as Ann was originally going to be a V1).

This is why CFM isn't considered the best company all in all. The best often debated one is AH Software.

Luka has one English vocal which they repeated for a second but just manually editing the first. "Soft" is the same voice. The English vocal range is small then the Japanese one and isn't the same as the V2 ones, which was identical to her Japanese range. Soft is the same library basically. This was one of the biggest "cheats" in V4x. The bugs are identical in both libraries in every way pretty much.

I'll correct you there. A Vocaloid can only do the languages each voicebank is built for. The example of Welsh is based on Phoneme manipulation, there are always bugs and so fourth left over. The bigger the differences the bigger the bugs. You can read our pages like English - Japanese for example. The result is LQ, though many refuse to accept that, for the simple reason Clarity is often regarded as the main source of quality among Vocaloids fans... I.e.... If I can understand it, its a good vocal (not always the case!). ITs the same in every case within Vocaloid, not just languages. Prima is HQ for opera but LQ for Blues, for example. Piko is good for mid-high male vocals, but lower down LQ. Though in cases like Low Piko, they like the vocal so they presume its HQ, and appeal tends to cloud peoples' opinions on the Vocaloids. Lets put it this way, its not a secret Act1 Kagamines is rubbish, but that doesn't stop people liking those vocals or songs with them in. ITs just how things go. We have a hell of a time explaining all this to fans who won't believe us at times, and often they are convinced their ears aren't lying. But... These fans will praise V2 Gumi results such as extremely high results, when Internet co admits those results are garbage... The human brain can piece together things even if their complete garbage and doesn't always know when something is bad, or when it does downplays it because of our own thoughts on the matter (pride, fantasy, ignorance, bias, etc). We're used to it...

One could compare it to the globe deniers of the Flat-Earth society who often believe the world is flat because their eyes deceive them into thinking the flat horizon means the earth is flat. Our perception of things, can often be incorrect completely. Its very easy to trick the human senses. Hell, if you smell chocolate while blind folded and are fed broccoli, your brain has to process while it can smell chocolate but taste something else. Welcome to being human. If your not convinced your brain can be that easily confused, just look up general "optical illusions", if your eyes can be tricked so can everything else. Even your sense of temperature can be tricked.

Vocaloid engines are released every 4 years to a predictable level. V4 was basically released early though and should have been out later that year. We should have been expecting V5 more later this year, as I myself expected Oct-Jan time. Why they have released it this early is unknown, but thats fair enough. The next Vocaloid engine release should be in 2022. Work normally begins on the 2nd year, so V6 will likely be in production serious in 2020. As I said, Vocaloid is predictable. The only thing I seriously got wrong myself was I somewhat thought it would break the cycle due to the extremely early release of V4, so I said it might come out same period last year. I did however say if it came out next year, it would be too late. So either way, it came out when expected, just not in the months I myself expected.