Does anyone prefer how Utau voices sound appose to Vocaloid?

[FeatheredFinch]

This is a very uncommon option but I feel a good UTAU voicebank can overpower a Vocaloid.

here's some reasons! to me the voice sounds crisper and clearer and more powerful (as long as it's high quality, otoed well and mixed well), UTAU english to me is way more understandable is less slurry even the vocaloids from english voice providers sound slurry to me. another reason is that it's free and can be modified and anyone can create a voicebank!

What do you all prefer?

 

[SynthJ]

I need a good mix of both I think XD Though I love the clarity of UTAU, Vocaloid does capture a more realistic essence of the English language- at least thats my opinion :D

 

[Kiyoteru]

I don't think it's uncommon at all to think that UTAU can surpass Vocaloid. The thing is, though they're both intended for singing synthesis, they're very different in terms of functionality and development. It's not entirely fair to compare them.
Though UTAU offers more potential/power/flexibility than Vocaloid, it's much less user friendly. Voicebank creation being open means there's no standard of quality, while with Vocaloid you can be guaranteed something at least as good as Sonika. UTAU has multiple engines. So, at this point in time, it's like "no duh" that some people have gotten very good at making the best of every aspect of it.

Then... there's situations like this: I once covered Mwk's remix of Love is War with Kiyoteru. I covered it again with an UTAU made out of his exVoice files. Even though the UTAU version was CV, and the mixing was lazier, it turned out sounding way more exciting and powerful. A good arguement for UTAU superiority, isn't it?

(Actually, the tuning on the UST was just really good. And Kiyoteru has soft high pitches, while a monopitch UTAU doesn't have that kind of limitation...)

 

[VocAddict]

I probably said this numerous times in numerous threads before but I can't really choose between them.

UTAU is less restrictive in its voice editing capabilites so it's so pretty easier to make it sound more expressive. Also, due to the numerous amount of engines the software, there's basically an engine for every voice type.

Vocaloid on the other hand, though it can sound bland in its unedited form, it can sound really really good when used correctly. There's the fact that Vocaloid has only one engine so if it doesn't like a voice and makes it sound like crap, you can't just swap out with a next resampler and fix it. It's more restrictive in some ways (can't manually choose pitches. That'll be awesome) but its synthesis capabilities are superior in my opinion.

As for language recognition, there were only a few voices I could understand properly in their raw state without lyrics from both software. That's mostly due to the accent that the banks are recorded in (I actually understand English banks from Japanese voice providers way easier than those from native speakers on both sides) and most times, they sound a bit forced.

All in all, I think it's kind of unfair to really pitch the both of them against each other. They both have their strengths and weakness, and they both can sound really good when pushed to the extreme.

 

[KNΞMΛTCS]

Vocaloid has a warmer sound to it, which I find pleasing to the ear. It doesn't struggle with complex blends like Utau does. On the other hand, sometimes that warmer sound is a disadvantage: there's times where Utau can really cut through a mix while Vocaloid just muddies it up. Vocaloid also has this issue, especially with male voices, where it sounds like the singer has a wad of paper in their throat or something.

Utau has a harsher tone, although less so with non-stock resamplers. It just sounds more 'robotic', especially since making it sound good is so hard. Its biggest advantage is its flexibility - you can change everything in Utau, which can be a boon when you're after a very specific sound or effect. However, this also gives Vocaloid an edge in ease of use - I can slap some lyrics right into Vocaloid and have usable output, while Utau requires me to convert those lyrics to sounds, align those sounds as needed, and probably do some pitch editing just to get what Vocaloid gave me right off the bat. Voicebanks are another 2-way street with Utau - you get a huge selection of voices to pick from, but at a cost of dealing with zero quality control.

I've said this earlier, but I like to compare Vocaloid to Windows and Utau to Linux. The first is fairly easy to use and gets the job done, but lacks in customization and flexibility. The latter is power user's heaven, with limitless extensibility but at a cost of user friendliness "just working".

 

[Kassandra]

I prefer how utau sounds

 

[Kiyoteru]

I like to compare Vocaloid to Windows and Utau to Linux.

I 100% agree with this comparison.

So, does this make Arpasing the Ubuntu?

 

[Arissa]

Lol this is the kinda thread I've been waiting for >=< I preferUtau over Vocaloid, period, voices being a factor, as well. As I had mentioned in a previous thread (I think) I think that Utau has more capabilities and can sound much more realistic/emotional than Vocaloid. Sure, Vocaloid is smooth, but to me can sound really unemotional, even the power voicebanks don't sound like they have much power. Vocaloid has this consistent sound that I think makes a lot of voicebanks sound similar, for example all the 'moe' type voices, and then there's Vocaloids meant to sound like Gumi apparently; Kokone, Yuzuki,, whereas Utau voices just sound more flexible and unique. There are some generic voices, but there are tons of banks in general. There are opera Utau, screamo Utau, smooth Utau, (purposely)robotic, etc.

To show what I'm talking about, especially when talking about power,, Darling Crazy Moon append, these words and this link

 

[VocAddict]

Vocaloid has this consistent sound that I think makes a lot of voicebanks sound similar, for example all the 'moe' type voices, and then there's Vocaloids meant to sound like Gumi apparently; Kokone, Yuzuki,, whereas Utau voices just sound more flexible and unique. There are some generic voices, but there are tons of banks in general. There are opera Utau, screamo Utau, smooth Utau, (purposely)robotic, etc.

That's basically what the market wants (V3 was the worse) and UTAU is free to create a bank so it's quite obvious it'll have tons more banks; whereas Vocaloid where companies have to pay money to make one.

They tend to stick to trends so they'll get guaranteed sales. That's why Vocaloid have so much similar banks (heck, Japan doesn't even like male voices that much which causes companies to make them quite rarely)

 

[Tomato Hentai]

VOCALOID is more easy to learn, whereas UTAU has more control over the way the end result sounds. Although UTAU can be harder to learn, once you know how to use it, then you might be able to get a nicer result. VOCALOID, while easy to learn, can be harder to get a nice sounding result.

I find that most English VOCALOID voicebanks have very strange and off-putting pronunciation quirks. They can sound too perfect and forced, or sometimes the exact opposite.
Cyber Diva's pronunciation sounds very forced, like if someone were trying to pronounce words by looking at their phonetic spelling and trying to match it 1:1. On the opposite (maybe?) side of the spectrum, you have DAINA who is supposed to be a Pop, Dance and Country vocal, but with her weird accent, she sounds almost like she's only supposed to be a Country vocal. Her pronunciation sounds so... Loose. Or maybe even too forced. I actually can't tell.
The (suspected) forcing of pronunciation kinda makes voices lose the qualities that make them unique, with the voice provider focusing primarily on how accurate their pronunciation is instead of them actually sounding like they're singing. This also seems to sometimes happen with certain VCCV and CVVC English UTAUᏚ sometimes.

The Japanese voicebanks are fine for the most part, we have quite a bit of variation with both VOCALOID and UTAU. I just wish we had one or two more Japanese Male VOCALOIDs.

because of what's going on in that other thread i forgot the other points i was trying to make

I think I might prefer UTAU over VOCALOID. It's a little more flexible with what you can do with it, and has some nicer voicebanks available for it.

 

[KNΞMΛTCS]

I 100% agree with this comparison.

So, does this make Arpasing the Ubuntu?

Close - it's Linux Mint.

 

[FeatheredFinch]

Close - it's Linux Mint.

What's the difference between english VCCV and Araparsing in terms of quality/pronunciation?

 

[Nohkara]

What's the difference between english VCCV and Araparsing in terms of quality/pronunciation?

I have used this two and I can say that this two has own strength and weaknesses.

VCCV:

Strengths
-most of the time it's just plug-and-play a premade VCCV UST, editing is not much required
-it sounds more smooth and more understandable than old CVVC Eng overall
-clear tutorials are available in CZ's YouTube

Weaknesses
-in my opinion unnecessary many "translation V C" oto settings
-"CC" of VCCV sounds easily in many VB... awkward
-there are some specific parts on VCCV that are very forceful to record for anybody: CV_CVC. (I think that this causes that "awkward odd accent " that many VCCV has)

Arpasing

Strength:
-super short English reclist with impressive results, much faster to record than VCCV
-because the reclist is mainly just lines of "real words", I think that makes beginners understand better how to record it.
-has the own website: basically, there's all that you need to know (link)

Weaknesses:
-no many tutorials available yet
-UST requires way more manual editing than VCCV, most of the time it's not "plug-and-play"
-pronouncing errors, mainly in vowels, more possible (I think this is cuz how reclist is written. E.g. Some people pronounces "doll" with [o] but some do that with [a] sound. The "ideal" pronouncing of "doll" is something between this two but you know: pronouncing variations)

 

[FeatheredFinch]

I have used this two and I can say that this two has own strength and weaknesses.

VCCV:

Strengths
-most of time it's just plug-and-play a premade VCCV UST, editing is not much required
-it sounds more smooth and more understandable than old CVVC Eng overall
-clear tutorials are available in CZ's YouTube

Weaknesses
-in my opinion unnecessary many "translation V C" oto settings
-"CC" of VCCV sounds easily in many VB... awkward
-there are some specific parts on VCCV that are very forceful to record for anybody: CV_CVC. (I think that this causes that "awkward odd accent " that many VCCV has)

Arpasing

Strength:
-super short English reclist with impressive results, much faster to record than VCCV
-because the reclist is mainly just lines of "real words", I think that makes beginners to understand better how to record it.

Weaknesses:
-no many tutorials available yet
-UST requires way more manual editing than VCCV, most of time it's not "plug-and-play"
-pronouncing errors, mainly in vowels, more possible (I think this is cuz how reclist is written. E.g. Some people pronounces "doll" with [o] but some does that with [a] sound. The "ideal" pronouncing of "doll" is something between this two but you know: pronouncing variations)

I find that Araparsing is less understandable sometimes but Vccv is sometimes a lot more awkward and sounds forceful, I tried to record my VCCV more naturally

 

[Sors]

I prefer UTAU over Vocaloid, mainly because i only had a Vocaloid trial, and thus didnt really learn how to use it well, and mostly used good vsqs and plug-and-play, while I learned how to work with UTAU, and now can produce fairy good music...well cover music.

 

[FeatheredFinch]

I don't think it's uncommon at all to think that UTAU can surpass Vocaloid. The thing is, though they're both intended for singing synthesis, they're very different in terms of functionality and development. It's not entirely fair to compare them.
Though UTAU offers more potential/power/flexibility than Vocaloid, it's much less user friendly. Voicebank creation being open means there's no standard of quality, while with Vocaloid you can be guaranteed something at least as good as Sonika. UTAU has multiple engines. So, at this point in time, it's like "no duh" that some people have gotten very good at making the best of every aspect of it.

Then... there's situations like this: I once covered Mwk's remix of Love is War with Kiyoteru. I covered it again with an UTAU made out of his exVoice files. Even though the UTAU version was CV, and the mixing was lazier, it turned out sounding way more exciting and powerful. A good arguement for UTAU superiority, isn't it?

(Actually, the tuning on the UST was just really good. And Kiyoteru has soft high pitches, while a monopitch UTAU doesn't have that kind of limitation...)

I just realised, don't use Sonika as a good quality example when she has skype bloops in her recording hahaha

 

[Damien]

I just realised, don't use Sonika as a good quality example when she has skype bloops in her recording hahaha

They should've done something about that

 

[FeatheredFinch]

I prefer UTAU over Vocaloid, mainly because i only had a Vocaloid trial, and thus didnt really learn how to use it well, and mostly used good vsqs and plug-and-play, while I learned how to work with UTAU, and now can produce fairy good music...well cover music.

I feel Utau could surpass in terms of english originals due to the ability to have a wide voice range.
[doublepost=1494423039][/doublepost]

They should've done something about that

I'm serious haha Myst talked about it im her recording livestream it was hilarious

 

[partial]

As someone who has Sonika I can confirm. Let me figure out what words or phonetics are affected.

 

[Damien]

A lot of common ones I believe @partialkakera