Project_VoXenuS

VocAddict

The Voice Within Us
Defender of Defoko
This is a thread to keep track of my main UTAU.

Primary (Product) Name: VoXenuS
- [The] Voice Within Us
- Deity of Emotion

Secondary (Character) Name: Hikaru Yami - 光ヤミ
- shining, glowing (Hikaru/光る/ヒカル)
- darkness (Yami/闇/ヤミ)

VoXenuS is a multilingual, multi-expression UTAU.

VoXenuS doesn't have a set design or a character, and is free to interpretation. The only fixed information is that he's male. Yami, on the other hand, has a (vast) backstory, but this information doesn't affect VoXenuS in any way (similar to unity-chan! and AKAZA). VoXenuS is the product while Yami is just a character using its voice.

Phase 1: World Series

Languages I'm positive about recording
> English
> Esperanto
> French
> Japanese
> Spanish

Languages I'm considering to record
> Dutch
> Italian
> Finnish
> German
> Greek
> Hebrew
> Korean
> Mandarin Chinese
> Portuguese
> Russian

Phase 2: Eternal Expression

Based on the original concept behind Yami, each bank will cover three tones in natural ascension of pitch: Dark, Shadow, and Light. (~F2 - F4)

Main Banks
VoXenuS_Eclipse (Natural)
VoXenuS_Soul/Strong (Power/Solid)
VoXenuS_Noct/Brittle* (Mellow/Gentle)
VoXenuS_Sky/Flight (Soft)
VoXenuS_Lune/Livid* (Whisper)

Considering - extra expression
(an extra pitch or two on the main bank)

VoXenuS_Sharp (Weak)
VoXenuS_Dull (Sharp)

Extensions
These banks would not have tone variations

VoXenuS_Growl
VoXenuS_Falsetto

* - name subject to change

  • The main banks would contain 7 pitches each (D2, F2, A2, C#3, F3, A3, and C#4)
  • The Growl bank would be recorded at E2 (subject to change)
  • The Falsetto bank would be recorded at G4 (subject to change), and may up to 3 different variants depending on the tone.
  • There's an unknown bank I haven't decided on yet.

I'm also thinking of making TTS banks to be released along side the bank for most of the languages.

> Natural
> Meak
> Aggressive
> Kind

Below are some samples of that can give you an idea of how VoXenuS will turn out in term of tone.

Japanese

English
https://clyp.it/31rvg0k3

Feel free to give any advice, suggestions, or just your general opinion.
 
Last edited:

Kiyoteru

UTAU Sensei
Supporter
Defender of Defoko
For the second phase, 5 banks times 3 tones results in 15 different tones (plus 6 more if you decide to include those other two!) That seems like a lot. Will you really be able to to distinguish them all clearly? Also, how many pitches do you think the banks will have, and what languages would the phase 2 expressions apply to? Multiplying all these numbers together, I'm kind of worried about how much there is.

But hey, at least there's Esperanto!
 
  • Like
Reactions: VocAddict

VocAddict

The Voice Within Us
Defender of Defoko
Thread starter
The distinguishing between expressions is something I've been deliberating about recently. I'm considering removing one of the banks (Mellow or Gentle) since over the past few weeks, my voice kind of merged the two of them together. But for the other four expressions, I believe I'll be able to tell them apart.

Sharp and Dull might end up being an extra pitch for each language and not really a tone for a bank, similar to the extensions, and it would be part for the main bank (Eclipse). Concerning the three tone issue, it covers my natural tone. So, dark to shadow to light. I'm trying to provide extra pitches for each of those tones just in case a user may want to achieve a certain expression at a certain pitch. I probably didn't phrase it properly.

Phase 2 would be for all languages. Each language will have an Eclipse bank by default and I'll try to do a minimum of 3 banks for each language. To be very honest, I see myself recording all these banks for Japanese and English at the moment. I'm not sure how much pitches each bank will take, mostly because I have no idea how to figure that out haha.
 
  • Like
Reactions: Kiyoteru

Alessandra

Ruko's Ruffians
Defender of Defoko
This is a thread to keep track of my main UTAU.

Primary (Product) Name: VoXenuS
- [The] Voice Within Us
- Deity of Emotion

Secondary (Character) Name: Hikaru Yami - 光るヤミ
- shining, glowing (Hikaru/光る/ヒカル)
- darkness (Yami/闇/ヤミ)

VoXenuS is a multilingual, multi-expression UTAU.

VoXenuS doesn't have a set design or a character, and is free to interpretation. The only fixed information is that he's male. Yami, on the other hand, has a (vast) backstory, but this information doesn't affect VoXenuS in any way (similar to unity-chan! and AKAZA). VoXenuS is the product while Yami is just a character using its voice.

Phase 1: World Series

Languages I'm positive about recording
> English
> Esperanto
> French
> Japanese
> Korean
> Spanish

Languages I'm considering to record
> Dutch
> Italian
> Finnish
> German
> Greek
> Hebrew
> Mandarin Chinese
> Portuguese
> Russian

Phase 2: Eternal Expression

Based on the original concept behind Yami, each bank will have three tones: Dark, Shadow, and Light. For example, for a Power bank, would have Power_Dark, Power_Shadow, and Power_Light.

Main Banks
VoXenuS_Eclipse (Natural)
VoXenuS_Soul/Strong (Power/Solid)
VoXenuS_Noct/Brittle* (Mellow)
VoXenuS_Sky/Flight (Soft)
VoXenuS_Lune/Livid* (Gentle)

Consdering - extra expression
VoXenuS_Sharp (Weak)
VoXenuS_Dull (Sharp)

Extensions
These banks would not have tone variations

VoXenuS_Growl
VoXenuS_Falsetto
VoXenuS_Whisper

* - name subject to change

I'm also thinking of making TTS banks to be released along side the bank for most of the languages.

> Natural
> Meak
> Aggressive
> Kind

Below are some samples of that can give you an idea of how VoXenuS will turn out in term of tone.


Feel free to give any advice, suggestions, or just your general opinion.
Really interesting project!
[doublepost=1501940464][/doublepost]
This is a thread to keep track of my main UTAU.

Primary (Product) Name: VoXenuS
- [The] Voice Within Us
- Deity of Emotion

Secondary (Character) Name: Hikaru Yami - 光るヤミ
- shining, glowing (Hikaru/光る/ヒカル)
- darkness (Yami/闇/ヤミ)

VoXenuS is a multilingual, multi-expression UTAU.

VoXenuS doesn't have a set design or a character, and is free to interpretation. The only fixed information is that he's male. Yami, on the other hand, has a (vast) backstory, but this information doesn't affect VoXenuS in any way (similar to unity-chan! and AKAZA). VoXenuS is the product while Yami is just a character using its voice.

Phase 1: World Series

Languages I'm positive about recording
> English
> Esperanto
> French
> Japanese
> Korean
> Spanish

Languages I'm considering to record
> Dutch
> Italian
> Finnish
> German
> Greek
> Hebrew
> Mandarin Chinese
> Portuguese
> Russian

Phase 2: Eternal Expression

Based on the original concept behind Yami, each bank will have three tones: Dark, Shadow, and Light. For example, for a Power bank, would have Power_Dark, Power_Shadow, and Power_Light.

Main Banks
VoXenuS_Eclipse (Natural)
VoXenuS_Soul/Strong (Power/Solid)
VoXenuS_Noct/Brittle* (Mellow)
VoXenuS_Sky/Flight (Soft)
VoXenuS_Lune/Livid* (Gentle)

Consdering - extra expression
VoXenuS_Sharp (Weak)
VoXenuS_Dull (Sharp)

Extensions
These banks would not have tone variations

VoXenuS_Growl
VoXenuS_Falsetto
VoXenuS_Whisper

* - name subject to change

I'm also thinking of making TTS banks to be released along side the bank for most of the languages.

> Natural
> Meak
> Aggressive
> Kind

Below are some samples of that can give you an idea of how VoXenuS will turn out in term of tone.


Feel free to give any advice, suggestions, or just your general opinion.
I know this can be in some part nrealistic, but I suggest you to not have accent (or try to not have), because you're doing a multilingual VB, so you have lots of languages, and you have to feel velievable. Sorry for this, becase it's not what I really think... I mean, I can't purely express it... for linguistic barrier ahah!
 

VocAddict

The Voice Within Us
Defender of Defoko
Thread starter
A quick update. The mic is finally here! A CAD GXL2400 USB. You can here me doing a horrible mic check here lol.

Recording should (hopefully) begin soon, along with tests for tone and pitch.
I know this can be in some part nrealistic, but I suggest you to not have accent (or try to not have), because you're doing a multilingual VB, so you have lots of languages, and you have to feel velievable. Sorry for this, becase it's not what I really think... I mean, I can't purely express it... for linguistic barrier ahah!
Ah, sorry for the late reply! I didn't get a notification for this.

Each language is being recorded separately as to avoid accent issues. There would be slight difference in accent, when switching languages though I hoping it would be minimal.
 

VocAddict

The Voice Within Us
Defender of Defoko
Thread starter
I didn't get to start recording yet, the house is never quiet lol. But I did manage to record a few vowel strings to compare with my previous bank. You can hear that small comparison here. Both banks are at the same pitch (F#2).

I finalised the pitches that I'll be recording for the banks.
  • The main banks would contain 6 pitches each (F2, A2, C#3, F3, A3, and C#4)
  • The Whisper bank would contain 7 pitches (D2, F2, A2, C#3, F3, A3, and C#4)
  • The Growl bank would be recorded at E2 (subject to change)
  • The Falsetto bank would be recorded at G4 (subject to change), and may up to 3 different variants depending on the tone.
  • There's an unknown bank I haven't decided on yet.
voxenus bank range.png

These pitches would apply for all languages, as long as the oto allows it. I haven't finalised what recording method I'll be using for Japanese do I created a poll. Feel free to vote.
 

VocAddict

The Voice Within Us
Defender of Defoko
Thread starter
So, I managed to get a get 30 minutes of silence so I recorded a quick test of my English reclist. You can check it out here. I didn't get to record everything needed and I'm sure that I mispronounced stuff in a few places, but I managed to make with what I had lol. I would upload the video of me making this test but it's like 100 minutes (323 MBs) with no audio for some reason so, nah lol.

The list is technically done but I want to make some changes before I record a full bank with it so it would be a while before any English bank gets released.
 
Last edited:

VocAddict

The Voice Within Us
Defender of Defoko
Thread starter
So, I recorded a CVVC bank in 10 minutes earlier today and decided to run a quick test with it lol. It's more of a test of how the mic sounds than a bank for release (though, if it's good enough, I might release it as a beta if I get around to other pitches). You can compare it to my previous bank from this clip.
 
  • Like
Reactions: Kiyoteru

VocAddict

The Voice Within Us
Defender of Defoko
Thread starter
So, I started to record VoXenuS_Eclipse bank earlier. I currently have F#2 finished but I'll mostly like record it again since there are some things I'm not satisfied with (and it feels more Noct than Eclipse as it is right now). Feel free to give some feedback and whatnot.


I've made some progress with the English reclist as well so there would be a test for that in a bit as well.
 
Last edited:

Anderson26

Ruko's Ruffians
Defender of Defoko
I love to see when people starting to making western utau project like viki, mocha and amane! They are really make me keep my attention to the future of the project!
 
  • Like
Reactions: VocAddict

VocAddict

The Voice Within Us
Defender of Defoko
Thread starter
School ends on the 3rd so I should have enough time to properly get some recording in before the new term starts. I'd get the Eclipse bank for Japanese out of the way and work my way from there.

I'll be merging the Noct and Lune together in terms of tones into one bank, which would be called Noct. Instead of just discarding the name Lune (I really, really like it), the whisper bank would be called that and that saves me from recording a whole 6 pitch bank lol. (will fix the image and the top post in a bit)

When I get the time, I'll do a quick recording test of the different tones and maybe a poll so you guys can help me decide what bank to record after Eclipse lol.
 

Similar threads