How do you make this and how long do you spend? It's amazing! My friend and I used to spend two weeks to try to let Renri sing in Chinese but failed because It sounds so strange. After that we tried English for two weeks too but also failed; it seems that she sings in Japanese not in English. Then we gave up.😭😭😭I think you did pretty well although it's a little strange. I will share it to my friend. ❤❤❤
@@hmzNO10 in short A lot Renri is a bad port of her R1 to R2 And R1 is already a bad port of 4 pitches of her utau Many combinations that are not japanese or when you try to do single consonants will produce background noise or create weird noise What I had to do, as you can see in the video, everytime Renri did that I did a curve in loudness so before a background noise start I reduce de volume After that, as I mix in FLSTUDIO i selected the option of "panneling volume" and everytime that noise came I reduce the volume to cut it For example I did "Time" as [ts a][i][m my] And Renri keep saying "Time-t" So reduce loudness to 0 in that last "t" sound so it wil be easy to delete in mixing I used the Whisper pitch which add random lip sounds and small breathes between every part In english you just have to play with phonemes For english "t" i used [ts] For english "r" i used [u][w] and the "u" and the "w" in phonemes was reduce to "30%" in the phonemes options For ending "r" i did the same with "r" For "L" i used the [r] phoneme and in phoneme options I set them in "150%" And thats pretty much what I did It was more fighting against Renri noise than anything so use Loudness a lot to reduce that
@@Sillypid Oh, I see. I often used tension, breathiness and voicing to solve these problem but avoided loudness, because I thought the default loudness was the best at most time. Maybe I need to change. And the English "r" pronounced by [u][w] is truly inspired me! I have a try and find it better than directly using [r] in some case! I may need to explore more to work with Renri. She is hard to tune.......Thank you very much for your help! ♥♥♥
@@Sillypid I have a try on the 1.4 version of SynthV and find it better than the latest version in timbre for Renri, but still bad than R1 version or UTAU version.......That's sad. Maybe Dreamtonics forgets to improve std.😭😭😭
@@hmzNO10 her STD was done just with Dreamtonics approval, the configs and all was done by Renri's team Is kinda messy but have good results I also Play with the parameters a lot
![[Sweet Ann] Perfect day / Cover](/img/n.gif)
Cool
How do you make this and how long do you spend? It's amazing! My friend and I used to spend two weeks to try to let Renri sing in Chinese but failed because It sounds so strange. After that we tried English for two weeks too but also failed; it seems that she sings in Japanese not in English. Then we gave up.😭😭😭I think you did pretty well although it's a little strange. I will share it to my friend. ❤❤❤
@@hmzNO10 in short
A lot
Renri is a bad port of her R1 to R2
And R1 is already a bad port of 4 pitches of her utau
Many combinations that are not japanese or when you try to do single consonants will produce background noise or create weird noise
What I had to do, as you can see in the video, everytime Renri did that I did a curve in loudness so before a background noise start I reduce de volume
After that, as I mix in FLSTUDIO i selected the option of "panneling volume" and everytime that noise came I reduce the volume to cut it
For example I did "Time" as
[ts a][i][m my]
And Renri keep saying "Time-t"
So reduce loudness to 0 in that last "t" sound so it wil be easy to delete in mixing
I used the Whisper pitch which add random lip sounds and small breathes between every part
In english you just have to play with phonemes
For english "t" i used [ts]
For english "r" i used [u][w] and the "u" and the "w" in phonemes was reduce to "30%" in the phonemes options
For ending "r" i did the same with "r"
For "L" i used the [r] phoneme and in phoneme options I set them in "150%"
And thats pretty much what I did
It was more fighting against Renri noise than anything so use Loudness a lot to reduce that
@@Sillypid Oh, I see. I often used tension, breathiness and voicing to solve these problem but avoided loudness, because I thought the default loudness was the best at most time. Maybe I need to change. And the English "r" pronounced by [u][w] is truly inspired me! I have a try and find it better than directly using [r] in some case! I may need to explore more to work with Renri. She is hard to tune.......Thank you very much for your help! ♥♥♥
@@hmzNO10 she is
Her voice quickly became the "No voice" sound of Synth V
Good luck
Don't give up
@@Sillypid I have a try on the 1.4 version of SynthV and find it better than the latest version in timbre for Renri, but still bad than R1 version or UTAU version.......That's sad. Maybe Dreamtonics forgets to improve std.😭😭😭
@@hmzNO10 her STD was done just with Dreamtonics approval, the configs and all was done by Renri's team
Is kinda messy but have good results
I also Play with the parameters a lot