Forums » Smalltalk » AMA: Vocaloid Expert, Furry, Fantasy Writer, Poly

I've been using vocal synthesizers, mainly UTAU, for a few years now and consider myself an expert in them. I currently manage an UTAU vocal called Jack Wicked, voiced by a fellow vocaloid lover TheTwelfthNote. I also use Vocaloid, SynthesizerV, Neutrino, CeVIO, Diffsinger and OpenUTAU. I also have DeepVocal, AISingers, Piapro Studio NT, Coeiroink, and AIVoice, but I don't use them as often.

I have been a furry since middle school, when I realized my gay awakening was Kyubi from Yo-Kai Watch. Once I got internet access, I joined the community and made my fursona, Jooters Blaccat. I am currently writing a fantasy novel about his universe.

I have also been polyamorous for about 2 years now, with both long distance and in person relationships.

AMA!

'sup bro! how far along are you with your novel?

Only a few pages in. I've been too busy for writing, unfortunately, but I hope to continue it... Eventually.
(This is also why I have yet to reply to our rp... Sorry about that.)

tbh from my experience the hardest part is even getting past the first page, once you roll it's easier to keep rolling? but that's a lot of rich talk coming from someone who hasn't worked on her books in years bwhahaha

either way, best of luck to you! hopefully you'll be able to get more work in soon.

(and I totally understand that, no worries at all! take your time)

I know I can obviously Google this, but I'd prefer the answer come from a person with personal experience: could you explain to me what vocaloids are? I'm aware of Hatsune Miku but I don't quite understand what she is, actually.

Also, good luck on the novel! My suggestion is to write an outline for the big plot points all the way to the ending before you dive in to the nitty gritty of writing it chapter by chapter. :>

Vocaloid is a vocal synthesis engine made by Yamaha made to be able to easily create synthetic vocals from a midi track. Essentially, you put in the notes you want the voice to hit along with the lyrics, and the program will create the singing for you. This is different to AI generation, as it's no different than using any other virtual instrument such as a DAW or midi keyboard.

In other words, they're instruments.

Vocaloid is also sometimes used to refer to singing vocal synthesizers, like how Coke is sometimes used to just mean all sodas, but other engines exist, such as Dreamtonic's Synthesizer V, which is famous for Yi Xi (Butcher Vanity) and Kasane Teto, or Ameya's UTAU, which is a free engine famous for Momo Momone (vocal from Nyan Cat) and... Also Kasane Teto.

Great explanation! I've been into Vocaloids for a long time, but I've never actually used one.
Do you find that the more expensive programs have better quality than the cheaper ones?

Quality? No, not typically.

I find that the free UTAU program gives out the best results most of the time if you know what you're doing, due to the nearly 20 years of people building onto it with plugins and engine improvements, although it can be quite complicated. Meanwhile, SynthV is insanely realistic and easy to use while around half the price of Vocaloid. And if you want something easy to use and free? Neutrino. It's Japanese input only, but insanely realistic.

In my opinion, most of the pricier programs nowadays (Vocaloid and CeVIO mainly) are mostly just cash grabs that rely on trendy characters and partnering with popular singers, voice actors or characters to advertise their products while offering little of anything new or quality.

I have yet to find a paid, non-SynthV vocal that is worth spending hundreds of dollars instead of just learning utau, setting up hifisampler, and downloading a descent to high quality vocal.

However, most fans of Vocal Synthesizers are more drawn to characters they already know, like Hatsune Miku, Kafu or SV Kasane Teto. So, if what you care about is more the character, then those programs are likely worth it, even if they aren't necessarily better quality wise to something like Mine Laru or Namine Ritsu running on hifisampler.

... Whew, sorry for the long rant just to say Miku is overpriced, lol.

No, I appreciate the in-depth response As a fan of Miku and co., I used to get fired up when people said UTAU was better, because I thought they were demeaning my favorites (and I love the weird varying unnatural pitchiness of Rin, Len, Gakupo, Kaito.... And Teto, who I thought was Vocaloid for a long time). But from a standpoint of variety and practicality and being a very active and well updated creative tool, I can better see the appeal of other programs, especially in the light of them all being the same tool and Vocaloid just being the most expensive one. I think the others just weren't as easy to get excited about, because it was harder to find information on them when I did research early on.
⁹

How fascinating! So it's like that program, Rockband, but with vocals? Pre-recorded samples used to create new pieces, and such?

Well... That's not entirely true. They all work rather similarly on a fundamental level (notes go in, voice comes out), they are all very different in their usage, operations and internals and have zero relation to each other in development. To say they're all the same tool is a little wrong because of how different they are in usage. They're all vocal synthesizers, but targeted to very different audiences and have very different functions.

Also, UTAU has not technically been officially updated in 12 years believe it or not! It's just the active community using the myriad of creation tools the engine offers. All the other major programs are frequently updated, Vocaloid included. The amount of different community made UTAU engines that are just on my computer (around 30-40) shows just how dedicated the community is to keeping it alive and high quality.

Something like that, but you don't use it with the direct samples. Each vocal is essentially a compressed file of thousands of little bits of audio that the program automatically applies filters to and strings together. You can then go in and edit the pitch curves and such. It's like any old midi keyboard, but fancier. This is also why it's so robotic.