Uploaded December 2023 | Updated September 2026, 5 hours ago
when I get a bad cough, there's only one way I can speak... and it's not a normal sick voice 😛
when I get a bad cough, there's only one way I can speak... and it's not a normal sick voice 😛


![[New vocal synth utan English test] Everything Stays from Adventure Time
Happy Thanksgiving my fellow Americans! I wanted to do something special for at least one holiday this year :P
New English test! Utan by Hataori-P / @hataori301
Everything Stays by Olivia Olson and Rebecca Sugar
We havent figured out the correct way to handle diphthongs yet, so thats why they might seem a little off. I used the flatten and contour intensity script, and thats why some of the zs went missing.
Utan is currently private software
Nicky B. might not have the best voice type for this song :P [New vocal synth utan English test] Everything Stays from Adventure Time](https://i.ytimg.com/vi/jlBgr81yzBk/mqdefault.jpg)

![[UTAUSnR] Wiki Wednesday Icon and Example Roundup #5 (41 - 50)
blog: https://utausnr.blogspot.com/
UTAU used:
skuidds wallows
梨音ナツカ / Nashine Natsuka
真歌ジョー / Joe Manaka
Toriga Happi / トリガーハッピー
Yochina Oto / 幼稚な音
薄歌 / Showoh Ryukas Hakka
美波うり / Minami Uri
スティーブ明日 / Steve Ashita
Inane Love / 伊奈音ラブ
来咲ライカ / Raika Kurusaki
I fixed the text settings to look more like the version done in VN! I cant remember if this method is slower or faster. [UTAUSnR] Wiki Wednesday Icon and Example Roundup #5 (41 - 50)](https://i.ytimg.com/vi/l2Sue7xQ-G0/mqdefault.jpg)
![Diffsinger Hanasu / Talkoid English Test featuring レウィ and Nicky B.
I did not expect Nicky to have that accent!
I recorded myself saying this and imported it into Praat. I labeled the phonemes the same way you label Diffsinger training data. I used my UST export script to export each note with the phoneme in brackets, so one note would be [s] and the next would be [ah]
I ran utalis on the output from Praat in OG utau. Saved that UST then imported it into OpenUtau.
Whatever you call the volume of each note, I didnt think it could be used by the Diffsinger renderer, but I was able to trace them. someone please tell me how to actually use parameters without manually drawing each part of the curve 🥲
outside of like rapping small sections and talkoids / HANASU, this method is extremely pointless. just use the utau at that point 😅 Diffsinger Hanasu / Talkoid English Test featuring レウィ and Nicky B.](https://i.ytimg.com/vi/l2nLfyScJM8/mqdefault.jpg)

![[UTAUSnR] Wiki Wednesday Icon and Example Roundup #17 (161 - 170)
UTAU Search and Rescue blog: https://utausnr.blogspot.com/
UTAUs used:
#161 - フックスケッキー / Keki Fökkusu
#162 - Dr. Leaf / ドクター・リーフ
#163 - Fukaina Shinpai / 心配 不快な
#164 - Chiki Kami
#165 - Huxley Pittal / ハクスリー ピタル
#166 - Minty Mascgrinn / マスクグリン ・ミンティー
#167 - D.SUE / ディ·ス
#168 - Tsuyuri Mochida / 持田つゆり
#169 - Kai Takahashi / かい たかはし
#170 - YA-01 ARTEMIS
There are some wonky sounding USTs used here and I have no idea how I made them or when I made them. Im pretty sure I have no access to them now, but Im still confused. [UTAUSnR] Wiki Wednesday Icon and Example Roundup #17 (161 - 170)](https://i.ytimg.com/vi/lB16MykFMII/mqdefault.jpg)

![[new singing synth utan test] え?あぁ、そう。 feat. 桃音モモ
I learned a lot today :) !
This can be improved further!
I felt like I was losing hearing in one ear and the tablet has really bad speakers, so the mix isnt great.
Momo has not been fully configured, so she can be improved further!
Hataori taught me a lot today about using Utan, but most important thing was actually how to edit pitch in Praat. did you know that in the previous version of this cover, one part was off key because the pitch ceiling was too low?! just changing the ceiling fixed it!
I edited the flatten and contour script I have always used to use intensity objects. That means that instead of sharing an mp3 for the sole purpose of using the intensity copy, I can just share the half megabyte intensity file!! (note: the audio file would still be required for the method involving utau unless Hataori redoes the script.)
what is utan? Utan is a new singing synth by @hataori301 which currently uses Praat as its GUI! The focus of Utan is voice copying. think Vocalo, but instead of a tuning mode... its the entire synth! There are no midis or notes or anything like that - everything comes from the textgrid for timing and the pitch data for... pitch. The engine has nearly no noise and its so clear and I am very much in love.
Eh, aa? sou by Chouchou-P
tuning copied from sm11088594
textgrid by me [new singing synth utan test] え?あぁ、そう。 feat. 桃音モモ](https://i.ytimg.com/vi/mYP6_OxsJGs/mqdefault.jpg)
