I recently discovered that Vocaloid 6 had come out a couple of years ago, and that it came with ten voice banks, all of which have not only Japanese phonemes, but also English and Chinese. Since I do largely classical music with choirs, the more voice banks I can get, the better. And since I'm usually working with Latin, this meant I could finally distinguish between R and L, and have real Fs and Vs. I downloaded the 30 day trial, and managed to overcome some of my initial issues, but some issues remained. Still, I loved some of the powerful voices like Shion and Asahi, so I took a risk and purchased Vocaloid 6. (I noticed also that the ZOLA voice banks had just come out remade for Vocaloid 6, but I'm hesitant to pull the trigger on them, though I use their V3 banks quite a lot.)
The method I'm using now is to make a track in the V4 Editor, where I find it also much easier to control dynamics and tempo. Then I import the .VSQX file into V6, where I can copy the Vocaloid track into a Vocaloid AI track, and then make any changes I want to the phonemes from Japanese to English or Chinese. So far so good.
One of the major problems that still remains for me is that I can't seem to get an extended even vibrato on all long notes, like I can in V4. I've looked at the subtitled Japanese tutorials, but what I've found in actual use is that some algorithm seems to decide what notes get vibrato, even when I select all notes and turn the vibrato depth up to 127. Some notes get little to no vibrato, while others have an unnaturally wide vibrato that I need to decrease. When the algorithm has decided that a note will not get vibrato, or will only get vibrato for the first quarter of the note, there seems to be nothing I can do to change it. In the V4 Editor, I can tell it to add a tasteful even vibrato to all notes a quarter note or longer, for example.
The other problem that remains is that there are enormous pitch bends at the beginning of many notes, and at the end of some notes. Again, an algorithm decides this, and I cannot seem to access it. When I select all notes and use the pitch tool to change the leading drift and following drift to 0, the result sounds very unnatural and robotic, quite unlike what I get from my V3 and V4 Vocaloids in the V4 Editor. So what I do is listen to the track, and only change those notes where the pitch bends are unacceptable.
I understand that all this is in an effort to sound more "human," and no doubt the V6 voice banks aren't intended for classical music (which is the main reason I haven't purchased V6 ZOLA yet). I like the flexibility of having phonemes from all three languages, and have had some success mixing V3, V4, and V6 Vocaloids in the V6 Editor, but the vibrato and pitch issues are still frustrating to me, and I wonder if I'm missing something. I'd appreciate any advice anyone might have.
Thanks in advance!
The method I'm using now is to make a track in the V4 Editor, where I find it also much easier to control dynamics and tempo. Then I import the .VSQX file into V6, where I can copy the Vocaloid track into a Vocaloid AI track, and then make any changes I want to the phonemes from Japanese to English or Chinese. So far so good.
One of the major problems that still remains for me is that I can't seem to get an extended even vibrato on all long notes, like I can in V4. I've looked at the subtitled Japanese tutorials, but what I've found in actual use is that some algorithm seems to decide what notes get vibrato, even when I select all notes and turn the vibrato depth up to 127. Some notes get little to no vibrato, while others have an unnaturally wide vibrato that I need to decrease. When the algorithm has decided that a note will not get vibrato, or will only get vibrato for the first quarter of the note, there seems to be nothing I can do to change it. In the V4 Editor, I can tell it to add a tasteful even vibrato to all notes a quarter note or longer, for example.
The other problem that remains is that there are enormous pitch bends at the beginning of many notes, and at the end of some notes. Again, an algorithm decides this, and I cannot seem to access it. When I select all notes and use the pitch tool to change the leading drift and following drift to 0, the result sounds very unnatural and robotic, quite unlike what I get from my V3 and V4 Vocaloids in the V4 Editor. So what I do is listen to the track, and only change those notes where the pitch bends are unacceptable.
I understand that all this is in an effort to sound more "human," and no doubt the V6 voice banks aren't intended for classical music (which is the main reason I haven't purchased V6 ZOLA yet). I like the flexibility of having phonemes from all three languages, and have had some success mixing V3, V4, and V6 Vocaloids in the V6 Editor, but the vibrato and pitch issues are still frustrating to me, and I wonder if I'm missing something. I'd appreciate any advice anyone might have.
Thanks in advance!