Head-Fi.org › Forums › Misc.-Category Forums › Members' Lounge (General Discussion) › The Touhou AND Vocaloid music thread!
New Posts  All Forums:Forum Nav:

The Touhou AND Vocaloid music thread! - Page 22

post #316 of 387
Can anyone suggest any non-touhou artists that are similar to Shibayan?
post #317 of 387
Thread Starter 

I wonder when is next HSP album coming up.

post #318 of 387
Quote:
Originally Posted by Ra97oR View Post

I wonder when is next HSP album coming up.

 

Some day™.

post #319 of 387

Wow, two big Vocaloid announcements today. First off, Hatsune Miku's English version has been confirmed to be coming out on her 6th birthday, August 31st. Download price 149 USD (tax excl.), available for both Windows and Mac and comes bundled with Studio One Artist digital audio workstation. Secondly her Japanese V3 version is being released on September 26th.

 

To contribute musically, here's an incredible GUMI song from 8#Prince.

 

 

Edit

Here is also a lovely Miku song which I may or may not have shared earlier in this thread. The artist is Miyazawa Moyoyo and the song is available as part of the Fix album which can be found on Bandcamp: http://quadrophenia.bandcamp.com/album/fix

 

 

Edit 2

Let's throw in an SF-A2 miki song and two featuring Avanna as well, 'cause Circus-P is so awesome. Frigid Hearts is especially mind-blowing. All three songs appears on the CUPCAKE -EP: http://circus-p.bandcamp.com/album/cupcake-ep

 

 


Edited by TJ Elite - 8/9/13 at 9:46am
post #320 of 387
Quote:
Originally Posted by TJ Elite View Post

Wow, two big Vocaloid announcements today. First off, Hatsune Miku's English version has been confirmed to be coming out on her 6th birthday, August 31st. Download price 149 USD (tax excl.), available for both Windows and Mac and comes bundled with Studio One Artist digital audio workstation. Secondly her Japanese V3 version is being released on September 26th.

...

 

post #321 of 387
Quote:
Originally Posted by Mad Max View Post

Quote:
Originally Posted by TJ Elite View Post

Wow, two big Vocaloid announcements today. First off, Hatsune Miku's English version has been confirmed to be coming out on her 6th birthday, August 31st. Download price 149 USD (tax excl.), available for both Windows and Mac and comes bundled with Studio One Artist digital audio workstation. Secondly her Japanese V3 version is being released on September 26th.

...

 

 

I know. What I'm still waiting the most myself is Luka Append, I love her so.

 


 

New HSP! Check it out!

 

 

Edit

New song from PTaka as well. What a nice day this is. Free download in the video description.

 


Edited by TJ Elite - 8/10/13 at 9:07am
post #322 of 387
Quote:
Originally Posted by TJ Elite View Post

 

I know. What I'm still waiting the most myself is Luka Append, I love her so.

 


 

New HSP! Check it out!

 

 

...

 

Thanks!  Hmm, not my kind of thing.

 

I would love a V3 or Append Luka as well.

post #323 of 387
Quote:
Originally Posted by Mad Max View Post

 

I would love a V3 or Append Luka as well.

 

Both are coming, eventually. The problem with her is that she's bilingual which makes the process much slower.


Edited by TJ Elite - 8/10/13 at 10:13am
post #324 of 387
Quote:
Originally Posted by TJ Elite View Post

 

Both are coming, eventually. The problem with her is that she's bilingual which makes the process much slower.

 

Why slower?

post #325 of 387
Quote:
Originally Posted by Mad Max View Post

Why slower?

 

Well Luka's Japanese and English both require their own voicebank so that's twice the amount of work. In addition her voice actress isn't a native English speaker so creating a good English voicebank is bound to be a slower process than creating a Japanese voicebank. And when we get into making an append with several different voice options life 'soft' etc., each of those requires their own voicebank as well; they aren't just digitally processed versions of the regular Luka. If she were to have for example the same number of voice options as Miku Append, which is six, that means Crypton has to make twelve new voicebanks for her, with half of them being in English. That's quite a lot of work. Even if they settle for less than six voice options for the append it's still quite an undertaking, especially if you are trying to deliver a quality product.

 

Edit

Oh and a Japanese voicebank requires samples for around 500 diphones per pitch; English needs 2,500. That's five times the amount of work. Poor Yū Asakawa…


Edited by TJ Elite - 8/10/13 at 12:26pm
post #326 of 387
Quote:
Originally Posted by TJ Elite View Post

 

Well Luka's Japanese and English both require their own voicebank so that's twice the amount of work. In addition her voice actress isn't a native English speaker so creating a good English voicebank is bound to be a slower process than creating a Japanese voicebank. And when we get into making an append with several different voice options life 'soft' etc., each of those requires their own voicebank as well; they aren't just digitally processed versions of the regular Luka. If she were to have for example the same number of voice options as Miku Append, which is six, that means Crypton has to make twelve new voicebanks for her, with half of them being in English. That's quite a lot of work. Even if they settle for less than six voice options for the append it's still quite an undertaking, especially if you are trying to deliver a quality product.

 

Edit

Oh and a Japanese voicebank requires samples for around 500 diphones per pitch; English needs 2,500. That's five times the amount of work. Poor Yū Asakawa…

 

...

Daaamn, so that's 500x88 and 2500x88 samples?  I've never looked into the program, but I suppose it is something similar to Kontact and such?

 

I guess that Asakawa-san probably may enjoy it if she is willing to sing an excerpt from "Just be Friends" when not even asked/expected(?) to.


Edited by Mad Max - 8/10/13 at 10:59pm
post #327 of 387
Quote:
Originally Posted by Mad Max View Post

...

Daaamn, so that's 500x88 and 2500x88 samples?  I've never looked into the program, but I suppose it is something similar to Kontact and such?

 

I guess that Asakawa-san probably may enjoy it if she is willing to sing an excerpt from "Just be Friends" when not even asked/expected(?) to.

 

I'm not sure where you got the number 88 (Number of notes in a certain number of octaves? I'm just guessing; my musical education wasn't all that good.), but my understanding is that in order to get natural sounding results, the Vocaloid software needs samples for three or four different pitch ranges; all other pitches within reason can be synthesized from those. I have no knowledge of how many pitches most Vocaloid soundbanks consist of in practice. That might vary case by case.

 

I'm not familiar with Kontact, but if you're interested in reading a bit about the software, I would recommend checking out the Wikipedia article for Vocaloid. It is quite technical, but basically it's based on using fragments consisting of different phonemes recorded at certain pitches to synthesize the same sound at a desired pitch and then using an algorithm to match the envelopes of joined sounds to get a smoother transition from one to the next. I'm sure there's more to it in practice, but I forget. Then there are of course various settings you can control like vibrato and so on, which are all based on some type of digital signal processing as well. That is at least what I remember gathering from reading the article about a year ago. It is quite possible I might have horribly misunderstood something, you never know. I actually think the basic working principle is quite simple really, but getting everything to work well is an entirely different matter and I'm really impressed with Yamaha's efforts so far. I really hope they will continued to further develop the software for it is a quite revolutionary tool for taking music into new directions.


Edited by TJ Elite - 8/11/13 at 2:56am
post #328 of 387
Quote:
Originally Posted by TJ Elite View Post

 

I'm not sure where you got the number 88 (Number of notes in a certain number of octaves? I'm just guessing; my musical education wasn't all that good.), but my understanding is that in order to get natural sounding results, the Vocaloid software needs samples for three or four different pitch ranges; all other pitches within reason can be synthesized from those. I have no knowledge of how many pitches most Vocaloid soundbanks consist of in practice. That might vary case by case.

 

I'm not familiar with Kontact, but if you're interested in reading a bit about the software, I would recommend checking out the Wikipedia article for Vocaloid. It is quite technical, but basically it's based on using fragments consisting of different phonemes recorded at certain pitches to synthesize the same sound at a desired pitch and then using an algorithm to match the envelopes of joined sounds to get a smoother transition from one to the next. I'm sure there's more to it in practice, but I forget. Then there are of course various settings you can control like vibrato and so on, which are all based on some type of digital signal processing as well. That is at least what I remember gathering from reading the article about a year ago. It is quite possible I might have horribly misunderstood something, you never know. I actually think the basic working principle is quite simple really, but getting everything to work well is an entirely different matter and I'm really impressed with Yamaha's efforts so far. I really hope they will continued to further develop the software for it is a quite revolutionary tool for taking music into new directions.

 

Well, 88 pitches like the 88 keys on a grand piano.

Apparently, none of the Vocaloids even have that much range anyway, haha.  Well, that's still a lot of work, in any event.  It looks like their differing pitch ranges may be dependent on the singer used to create the samples.

post #329 of 387
Quote:
Originally Posted by Mad Max View Post

 

Well, 88 pitches like the 88 keys on a grand piano.

Apparently, none of the Vocaloids even have that much range anyway, haha.  Well, that's still a lot of work, in any event.  It looks like their differing pitch ranges may be dependent on the singer used to create the samples.

 

That makes sense, although I would never have guessed.

 

It makes sense that the optimal vocal range of a Vocaloid would be largely dependent on person voicing them. I haven't used the program myself, but I believe it allows you to go quite high and low with the vocals but the results aren't very good if you go too far outside the recommended vocal range. That is my understanding anyway.

 


 

New song from Haloweak who is part of a Chinese electronic Vocaloid circle called Vanguard Sound. They are planning an international digital release of their next album on iTunes, Amazon and Beatport.

 

 

Edit

I just read on Facebook that Crypton might be releasing V3 MEIKO later this year. That's good because I like her.


Edited by TJ Elite - 8/12/13 at 8:09am
post #330 of 387
Quote:
Originally Posted by Hente View Post

Can anyone suggest any non-touhou artists that are similar to Shibayan?

Justice?

 

New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › Members' Lounge (General Discussion) › The Touhou AND Vocaloid music thread!