5 Newb Questions

SOHBlue

Master of Doom
Defender of Defoko
Hello, there! I've been on this site about a week now after just starting to work with UTAU and have learned a lot from this helpful community. I really do appreciate it and hope I haven't annoyed anybody too much. Rather than ask five questions in the chat box like my other individual questions I'd thought I post them here. Forgive me if they sound stupid as I'm still quite new (though not totally new, I know how to get voicebanks and whatnot) to this and know nothing about the Japanese language. On with it then:

1) Ls. (Li,La,Le) In a few voicebanks I'm working with I see there are no Ls like there are with other voice banks. These are the ones that either are or were originally in Japanese though. So I'm guessing that maybe they just used another letter to represent the sound. Like how I in Japanese is pronounced like E in English. I might just be being stupid, but if you could correct me if I'm wrong that'd be nice. ^.^

2) Kana and Hiragana. So after some tinkering, I got my English keyboard to type in Hiragana and listen to the Hiragana encoded voicebanks I'm working with. Hooray! But, when I tried it with others, nothing. I found that they were aliased in Katakana, but when I switched the keyboard and typed in Kana-nothing still. Why not?

3) Flags. So I already asked someone to help explain them a bit better, but I still didn't quite get it. If you could give me a simplified explanation, I'd appreciate it.

4) Plug ins. I know what they are, and I have some such as NoBreath, Humanizer and the ry one by Bizz. However, though I click on them in UTAU and they say they are working, when I play the parts I want breaths in or to sound more human- nothing. Hmmm...

5) Multiple folder voicebanks. In particular, Yami's English voicebank has multiple folders and a bunch of different sounds in each. One person told me I shouldn't have to get them all in one folder as it should put in all of the voicebank sounds, but this isn't the case. Only the sounds from the folder I got the first sound from goes in, not the whole bank. Yami has folders labled, _ CV, up arrow, down arrow, CC-, CCV, CV, CV_CVC, CVC_CV, V, VC and VCC...Wow. So yeah, I'm confused with his voicebank...

Anyways, thank you for taking the time to read all of my crap and I hope you can help me sort through it.
 
  • Like
Reactions: CRTブラック

LunarConstruct

Ritsu's Renegades
Defender of Defoko
Hi! i'm gonna see if i can help ya here -

1) The Japanese language doesn't have the "l" sound at all. Most Japanese voicebanks use their "r"s instead, because it is the closest sound. Some reclists include "l", but not all of them do so you can't always count on it.

2) I'm guessing why it's not working is because the voicebank you're trying to use doesn't have hiragana encoding or aliases. Check the otos to see whether it uses romaji or hiragana to be sure!

3) There are a toooon of flags that are used in the voicebank selection screen to alter the rendering/voice. Some are pre-render, some are post-render, and they can do things like make the voice breathier, deeper, etc.

4) I can't help you on this one, I've never used these plugins ^^;

5) Yami's English uses VCCV (and by the arrows I'll assume they are are also multipitch). VCCV is seperated into folders for easier recording and otoing - and multipitch just means the voicebank is recorded in another pitch as well as the base pitch. You shouldn't have to do anything to the voicebank to use them. Everything is already set up for you, so just select the main folder that hosts all the smaller folders, and you should be good!

I hope this helps~
 
  • Like
Reactions: SOHBlue

SOHBlue

Master of Doom
Defender of Defoko
Thread starter
Hi! i'm gonna see if i can help ya here -

1) The Japanese language doesn't have the "l" sound at all. Most Japanese voicebanks use their "r"s instead, because it is the closest sound. Some reclists include "l", but not all of them do so you can't always count on it.

2) I'm guessing why it's not working is because the voicebank you're trying to use doesn't have hiragana encoding or aliases. Check the otos to see whether it uses romaji or hiragana to be sure!

3) There are a toooon of flags that are used in the voicebank selection screen to alter the rendering/voice. Some are pre-render, some are post-render, and they can do things like make the voice breathier, deeper, etc.

4) I can't help you on this one, I've never used these plugins ^^;

5) Yami's English uses VCCV (and by the arrows I'll assume they are are also multipitch). VCCV is seperated into folders for easier recording and otoing - and multipitch just means the voicebank is recorded in another pitch as well as the base pitch. You shouldn't have to do anything to the voicebank to use them. Everything is already set up for you, so just select the main folder that hosts all the smaller folders, and you should be good!

I hope this helps~

It does, thank you! ^.^
[doublepost=1458672261][/doublepost]Okay, I have worked with the Kana voicebanks I mentioned and they did have some romaji aliases as well as hiragana, but they won't work either way. Their info and icon is up, so I wouldn't think the voicebank didn't load properly, but I suppose I should double check...
 
  • Like
Reactions: CRTブラック

na4a4a

Outwardly Opinionated and Harshly Critical
Supporter
Defender of Defoko
Hello, there! I've been on this site about a week now after just starting to work with UTAU and have learned a lot from this helpful community. I really do appreciate it and hope I haven't annoyed anybody too much. Rather than ask five questions in the chat box like my other individual questions I'd thought I post them here. Forgive me if they sound stupid as I'm still quite new (though not totally new, I know how to get voicebanks and whatnot) to this and know nothing about the Japanese language. On with it then:

1) Ls. (Li,La,Le) In a few voicebanks I'm working with I see there are no Ls like there are with other voice banks. These are the ones that either are or were originally in Japanese though. So I'm guessing that maybe they just used another letter to represent the sound. Like how I in Japanese is pronounced like E in English. I might just be being stupid, but if you could correct me if I'm wrong that'd be nice. ^.^

2) Kana and Hiragana. So after some tinkering, I got my English keyboard to type in Hiragana and listen to the Hiragana encoded voicebanks I'm working with. Hooray! But, when I tried it with others, nothing. I found that they were aliased in Katakana, but when I switched the keyboard and typed in Kana-nothing still. Why not?

3) Flags. So I already asked someone to help explain them a bit better, but I still didn't quite get it. If you could give me a simplified explanation, I'd appreciate it.

4) Plug ins. I know what they are, and I have some such as NoBreath, Humanizer and the ry one by Bizz. However, though I click on them in UTAU and they say they are working, when I play the parts I want breaths in or to sound more human- nothing. Hmmm...

5) Multiple folder voicebanks. In particular, Yami's English voicebank has multiple folders and a bunch of different sounds in each. One person told me I shouldn't have to get them all in one folder as it should put in all of the voicebank sounds, but this isn't the case. Only the sounds from the folder I got the first sound from goes in, not the whole bank. Yami has folders labled, _ CV, up arrow, down arrow, CC-, CCV, CV, CV_CVC, CVC_CV, V, VC and VCC...Wow. So yeah, I'm confused with his voicebank...

Anyways, thank you for taking the time to read all of my crap and I hope you can help me sort through it.

1) Japanese doesn't have an L sound in the same context of english. it is just another variant of the japanese r sound. Where instead of a tongue flap it's a very sharp, fast, and light l-esk sound. It's made with the tongue in an entirely different place too.

2) That shouldn't of taken tinkering, are you sure you changed your locale? It should be set and ready. There should be not voicebanks aliases in katakana, hiragana is the standard way of inputting sound. We need more info on this,

3) There are A LOT of flags and they change depending on which resampler you use.
Here are the more important ones for the default resampler and fresamp:
g: gender flag, default is g0, negative values make the voice higher/squeakier, positive values make the voice lower and deeper. It's range is g-100 to g100
B: breath, default B50, just how much "hiss" and breath is in the voice. it's range is B0 to B100
G: regenerate the frq file for your sample, don't confuse with g.
F: Formant filter, controls the prevention of the "squirrel" effect when shifting pitches. You know how when you pitch your voice up in audacity it gets overly squeaky? this is a control for that. In most cases leave it be but for fresamp try setting it to "F0.5"

4) let me get you an image

5) That's VCCV. It's the new way of doing English that's all the range and is indeed very confusing in it's organization. folders don't matter as long as they only go one folder deep (utau won't see folders deeper than this). You could have none, you could have 400, as long as the bank works.
[doublepost=1458673296][/doublepost]
bV1oubM.png

Do this in the utau options. This will allow older plugins to work. Because in the current version of utau they just don't by default.
 
  • Like
Reactions: SOHBlue

SOHBlue

Master of Doom
Defender of Defoko
Thread starter
1) Japanese doesn't have an L sound in the same context of english. it is just another variant of the japanese r sound. Where instead of a tongue flap it's a very sharp, fast, and light l-esk sound. It's made with the tongue in an entirely different place too.

2) That shouldn't of taken tinkering, are you sure you changed your locale? It should be set and ready. There should be not voicebanks aliases in katakana, hiragana is the standard way of inputting sound. We need more info on this,

3) There are A LOT of flags and they change depending on which resampler you use.
Here are the more important ones for the default resampler and fresamp:
g: gender flag, default is g0, negative values make the voice higher/squeakier, positive values make the voice lower and deeper. It's range is g-100 to g100
B: breath, default B50, just how much "hiss" and breath is in the voice. it's range is B0 to B100
G: regenerate the frq file for your sample, don't confuse with g.
F: Formant filter, controls the prevention of the "squirrel" effect when shifting pitches. You know how when you pitch your voice up in audacity it gets overly squeaky? this is a control for that. In most cases leave it be but for fresamp try setting it to "F0.5"

4) let me get you an image

5) That's VCCV. It's the new way of doing English that's all the range and is indeed very confusing in it's organization. folders don't matter as long as they only go one folder deep (utau won't see folders deeper than this). You could have none, you could have 400, as long as the bank works.
I really appreciate your reply. I will keep it for reference, I am sure I have my locale set and good. My other Hiragana voicebanks like Ruko are working flawlessly. It is Ryouji Kizuato and Linako Sonone that will not work with me.
 
  • Like
Reactions: CRTブラック

AsheNeff

Teto's Territory
im also a noob but since no one is mentioning it

5. When you have samples in separate folders; just select THE OTO in the main folder; as the folder itself is not selectable. I had the same problem before with CZloid. if you select a sound file from one of the sample folders it will think that folder is the root for the VB. You need a sound file or the oto file from the main folder; and then it should work.
 
  • Like
Reactions: SOHBlue

SOHBlue

Master of Doom
Defender of Defoko
Thread starter
1) Japanese doesn't have an L sound in the same context of english. it is just another variant of the japanese r sound. Where instead of a tongue flap it's a very sharp, fast, and light l-esk sound. It's made with the tongue in an entirely different place too.

2) That shouldn't of taken tinkering, are you sure you changed your locale? It should be set and ready. There should be not voicebanks aliases in katakana, hiragana is the standard way of inputting sound. We need more info on this,

3) There are A LOT of flags and they change depending on which resampler you use.
Here are the more important ones for the default resampler and fresamp:
g: gender flag, default is g0, negative values make the voice higher/squeakier, positive values make the voice lower and deeper. It's range is g-100 to g100
B: breath, default B50, just how much "hiss" and breath is in the voice. it's range is B0 to B100
G: regenerate the frq file for your sample, don't confuse with g.
F: Formant filter, controls the prevention of the "squirrel" effect when shifting pitches. You know how when you pitch your voice up in audacity it gets overly squeaky? this is a control for that. In most cases leave it be but for fresamp try setting it to "F0.5"

4) let me get you an image

5) That's VCCV. It's the new way of doing English that's all the range and is indeed very confusing in it's organization. folders don't matter as long as they only go one folder deep (utau won't see folders deeper than this). You could have none, you could have 400, as long as the bank works.
[doublepost=1458673296][/doublepost]
bV1oubM.png

Do this in the utau options. This will allow older plugins to work. Because in the current version of utau they just don't by default.
Alright, thank you!
[doublepost=1458673595][/doublepost]
im also a noob but since no one is mentioning it

5. When you have samples in separate folders; just select THE OTO in the main folder; as the folder itself is not selectable. I had the same problem before with CZloid. if you select a sound file from one of the sample folders it will think that folder is the root for the VB. You need a sound file or the oto file from the main folder; and then it should work.
Got it and on it, thanks guys! ^.^
 
  • Like
Reactions: CRTブラック
Similar threads

Similar threads