/leek/ - Vocaloid Lounge

>>227
I prefer TuneLab. Wish the UTAU compatibility was better because compared to OpenUTAU it's extremely laggy.

Do you love me too anon?

What resampler do you guys use and why is it not doppeltler?

I’m having some problems with openutau exports. Anyone had this?
>.wav output seems out of sync with the BGM
>Either it’s rendering too fast or some notes are rendering shorter than expected - audio output is faster than BGM
>Checked tempo across project and music program, they match
Somebody listed this on the github but it got closed with no updates. Maybe it’s user error and I’m missing something silly?
>>2153
I’ve found doppleter sounds good on high-pitched voices but makes low-pitched voices sound boxy. Moresampler is better in some cases

>>4465
I would just render the vocals first and add the bgm later via DAW like Fl Studio.

>>2153
>What resampler do you guys use and why is it not doppeltler?
70% of resamplers hate my UTAU and Doppeltler is no exception.
UST is mostly untuned because I'm too lazy to tune it right now
Straycat-rs (works): https://files.catbox.moe/4y0a55.wav
Doppeltler: https://files.catbox.moe/cdllsd.wav
Honestly I usually refuse to use any world-based resamps that aren't Straycat.

>>4466
I thought I was having technical issues, but looking at the instrumental in the DAW, I think it's not the software but the song. Although this song is labelled as 128 BPM it just kind of casually fluctuates down to 127 BPM and back. Very cool, very normal
>>4467
Agreed that WORLD usually sounds bad.

>>4469
I know it's unrelated but if you render the midi at a BPM like 127.555 then OpenUTAU will read the bpm as 127, that's one thing I learned when making covers. Tunelab doesn't have this problem.

My top 3 resamplers are Hifisampler (it makes any utau sound like a vocaloid), Moresampler and Doppeltler.

>>4471
If hifisampler gets a linux release I’ll finally have no reason to boot into windows for utau. That would be great.

Speaking of HiFiSampler. Someone on Twitter made a portable version of 0.0.5 which doesn't require installation. It's slightly slower than the original but it works.
https://x.com/crazy_toho/status/1912893357418053985
https://10.gigafile.nu/0617-s96be0a87906d969409248bd8a4070231

>>4469
Yeah, with WORLD you have to be extremely careful. I like Straycat (though my favorite is Hifisamp, which uses AI instead of WORLD) as far as WORLD based goes. It exports a bit buzzy but the lo-fi quality works well with some songs.
>>4470
I don't get the appeal of Tunelab honestly. I can't wrap my head around the basics and am Windows Only with my vsynth programs though.
>>4471
Hifisamp is really good but having to install to Win32 and do copious amounts of Python setup only for most UTAUs to spit out errors like no tomorrow sucks. It also has zero flags so vocals always sound the same. THAT SAID, it is overall my favorite resampler. The quality is incredible
Whether Moresampler works at all with any vocals is a major gamble. It doesn't work with my UTAU, which is CVVC, but with most VCV ones it sounds incredible. The exception are some baritone or bass voices, which work better with TIPS. TIPS is also pretty buggy, but in the opposite direction. It doesn't work with tenor voices or higher.

>>4476
>Hifisamp is really good but having to install to Win32 and do copious amounts of Python setup

See >>4474

>>4476
I used Vocaloid4FE for my covers in the 2010's but ever since Tunelab released it's been my go-to software for tuning Vocaloids and the pitch drawing feature is basically a must have. It's essentially V6 but on steroids.

>>4476
Isn’t tips the one that sounds great on soft/whisper banks? I mean it makes all the voices in that bucket sound kind of the same but it’s not a bad sound.

>>4477
Thanks anon, I meant to reply and say thank you and that I'd try it out but I forgot.
>>4478
Thanks anon. I think the best Vocaloid program to tune with was, ironically, V2. Nowadays I tune in OpenUTAU and port it to V4 using UtaFormatix. Sometimes it needs to be tweaked but usually the tuning is cross compatible.
>>4479
Yeah, TIPS is pretty good for that. It tends to bring out all the qualities of the voicebank with higher pitched vocals having increased background or engine noise (depends on the bank) so it isn't usually used on vocals like Teto or Momo

Funniest shit i've seen today

>>4481
I can't believe /jp/ finally put out content.

>>4481
I'm a vocaboomer too but lol

>>4480
I personally can't get used to old Vocaloid's UI these days. I'm a vocaboomer myself and grew up making shitty covers on Pocaloid so i technically do know how to use it but in retrospective it's so shit compared to what the community has made. Vocaloid6 does look promising but i'm not paying for that overpriced crap that sounds like Cevio but worse.

>>4485
That's understandable. Everything feels primitive in comparison to SV1 to me, but I prefer the old concat sound most of the time so I cycle editors. Hopefully when I can figure out where to find TLX extensions I can try TuneLab out fully.

>>4481
To be fair i get where anon is coming from. The first result for Tunelab is literally that 90's looking piano tuning software.

So there's a mobile version of OU now

Any clue what this "UtauV" i've been seeing around is? Seems like it's a fork of OpenUTAU?

>>24749
Fork of OpenUtau packed with some extra features, yeah. Creator recently released an update allowing for XSY with UTAU banks so that's why it's getting attention I think

>>2475
XSY support in UTAU sounds cool

>>24749
they're adding collab mode to it apparently

>>688
Who

>>27641
cant help but feel a bit bad for deepvocal sometimes lol