- Thread starter TheJoshMorin
- Start date
![[K] of Honor](/data/medal/1_1406930084l.jpg)
UTAU itself has not received any software updates, though the developer has released three new resamplers: doppeltler, f2resamp, and phaavoco (with two As).
If you primarily intend to quickly record a voicebank, automatically generate the OTO, not make any adjustments, then your choice of reclists is appropriate. These are voicebanks that you can make for your own use to quickly prototype ideas, especially in the case of Arpasing. Without further refinement, I wouldn't advise releasing them to the public, unless they need to be released as a sort of beta version that people can use while you're still working on finishing them or making new voicebanks.
If you're willing to put in the work to fully adjust the OTO by hand, then the most efficient Japanese reclist would be CVVC. An efficiently written reclist could be less than 50 strings total, and the amount of OTOing should only be twice as much as a CV voicebank.
If you primarily intend to quickly record a voicebank, automatically generate the OTO, not make any adjustments, then your choice of reclists is appropriate. These are voicebanks that you can make for your own use to quickly prototype ideas, especially in the case of Arpasing. Without further refinement, I wouldn't advise releasing them to the public, unless they need to be released as a sort of beta version that people can use while you're still working on finishing them or making new voicebanks.
If you're willing to put in the work to fully adjust the OTO by hand, then the most efficient Japanese reclist would be CVVC. An efficiently written reclist could be less than 50 strings total, and the amount of OTOing should only be twice as much as a CV voicebank.
Hewoo!
Edit1: Sample = sound that you recorded.
I'm not really that good with this kind of thing, but I think I can help you with smth. Starting with an answer to your deepvocal question Idk for sure, but I'll be honest I think yes although I recommend you to start with utau (Just bc you have more content and bc I preffer it).
First you have to decide what voicebank do you want to record, if it is a CV, VCV, CVVC, etc. As you said Arpasing is actually considered by the community as the best english method, but it doesn't mean that others recording styles(methods) are bad too. Tbh Arpasing can be really hard to do if you don't have much experience with utau. Considering that CVVC and VCCV (A method created exclusively for English vbs that you can found more information here: http://utau.us/vccv.html) are really good recording methods and they have both languages support you just have to search their reclists.
What I really recommend you to do is search a reclist that support both languages to the method that you chose, but if you don't find it you can use two reclists (One for English and another one for Japanese) to record your utau. You can use preffixes to differentiate the languages. Preffixes are characters that we write after or before the sample's character to differentiate same samples sound for the utau. For example: If you record "Ra" in C5 but all your others samples was recorded in C4 pitch (Even a "Ra" sample recorded in C4). But this can turn your vb hard to use and confusing depending on what you want to do with your utau.
But there's a better solution to this. Here where I live an utau creator made a method to record portuguese voicebanks using a language named by X-Sampa. X-Sampa is a language created to the phonetical international alphabet, Idk much about it, but if you can't use X-Sampa you can name your utau's sample with the International Phonetical Alphabet (IPA) to the English language and name japanese samples with the IPA/X-Sampa symbols. Obviously that the sounds that both languages share between them can be writhe in a normal way without using any symbol, doing this you will help your utau's users to have less trouble when they're going to do an utau ust.
I hope I could help you, if not I can explain to you better latter! ^^ Thank you!
See ya~
Edit1: Sample = sound that you recorded.
I'm not really that good with this kind of thing, but I think I can help you with smth. Starting with an answer to your deepvocal question Idk for sure, but I'll be honest I think yes although I recommend you to start with utau (Just bc you have more content and bc I preffer it).
First you have to decide what voicebank do you want to record, if it is a CV, VCV, CVVC, etc. As you said Arpasing is actually considered by the community as the best english method, but it doesn't mean that others recording styles(methods) are bad too. Tbh Arpasing can be really hard to do if you don't have much experience with utau. Considering that CVVC and VCCV (A method created exclusively for English vbs that you can found more information here: http://utau.us/vccv.html) are really good recording methods and they have both languages support you just have to search their reclists.
What I really recommend you to do is search a reclist that support both languages to the method that you chose, but if you don't find it you can use two reclists (One for English and another one for Japanese) to record your utau. You can use preffixes to differentiate the languages. Preffixes are characters that we write after or before the sample's character to differentiate same samples sound for the utau. For example: If you record "Ra" in C5 but all your others samples was recorded in C4 pitch (Even a "Ra" sample recorded in C4). But this can turn your vb hard to use and confusing depending on what you want to do with your utau.
But there's a better solution to this. Here where I live an utau creator made a method to record portuguese voicebanks using a language named by X-Sampa. X-Sampa is a language created to the phonetical international alphabet, Idk much about it, but if you can't use X-Sampa you can name your utau's sample with the International Phonetical Alphabet (IPA) to the English language and name japanese samples with the IPA/X-Sampa symbols. Obviously that the sounds that both languages share between them can be writhe in a normal way without using any symbol, doing this you will help your utau's users to have less trouble when they're going to do an utau ust.
I hope I could help you, if not I can explain to you better latter! ^^ Thank you!
See ya~
Big_B, from OP's self-introduction post, they're a returning UTAU user who already has experience. This post was not a question about how to record a voicebank, but rather, wanting to know which of the options would fit their particular situation of wanting to record a voicebank quickly and generate a Moresampler OTO. They are already well aware of the various methods, so your essay was reiterating information that was already known. You may have also missed the part where they already specified wanting to do separate English and Japanese voicebanks- in which case, a single combined reclist is not what they are looking for.
Zillaloid, here's where the latest version of the resamplers can be downloaded.
doppeltler: http://utau2008.xrea.jp/2020/engine/doppeltler009.zip
f2resamp: http://utau2008.xrea.jp/2020/engine/f2resamp005.zip
phaavoco: http://utau2008.xrea.jp/2020/engine/phaavoco001.zip
There's also a 64 bit wavtool: http://utau2008.xrea.jp/2020/engine/wavtool64.zip
Older versions of doppeltler, as well as information about the changes, are on the website: http://utau2008.xrea.jp/2020/engine/
And you can also read Ameya's tweets for information about these resamplers: https://twitter.com/ameyaP_
Zillaloid, here's where the latest version of the resamplers can be downloaded.
doppeltler: http://utau2008.xrea.jp/2020/engine/doppeltler009.zip
f2resamp: http://utau2008.xrea.jp/2020/engine/f2resamp005.zip
phaavoco: http://utau2008.xrea.jp/2020/engine/phaavoco001.zip
There's also a 64 bit wavtool: http://utau2008.xrea.jp/2020/engine/wavtool64.zip
Older versions of doppeltler, as well as information about the changes, are on the website: http://utau2008.xrea.jp/2020/engine/
And you can also read Ameya's tweets for information about these resamplers: https://twitter.com/ameyaP_