#writing this sapped so much energy from me idc how late this is posted

[utau-bowl]

Looking at all the Voicebanks on the Arpasing website

This is a project that took a good amount of time to write and publish, so I’m glad I’m finally here! For context, there is an unofficial Arpasing website with reclists, plugins, and USTs to use. There’s also a list of avilable Arpasing voicebanks to try out, and here’s where I had the idea to download and talk about-all of them. These will generally be pretty short, but I hope they’re at least somewhat informative. Many of these voicebanks are in beta/unfinished to a degree, so keep that in mind. Since a lot of these are old beta voices, if you don’t want your VB on here, just send me a message and I’ll delete it.

All of these voices can be downloaded at the Voicebank Directory on the Arpasing site. 

A few of the links on the site didn’t work, so sorry Cherry, Daichi Hayashi, and Armani. I’m also not including the Arpasing prototype VB since it’s just recordings and has no oto. For people who don’t know much about arpasing: 0.1.0 is the original version of Arpasing with 219 samples, and 0.2.0 is an updated later version made by Kanru Hua that has 334 samples. I also bring up the alveolar tap a little bit. The alveolar tap is a consonant common in american English (it’s the “tt” sound in butter, better, and letter). Since Arpasing was optimized for American English, it’s surprising that most of the reclists for it don’t include the alveolar tap! So I might point that out when I see it added into a vb.

Aero-gen Shinsu by Wolfe Nebula

※Reclist: 0.1.0 ※Number of pitches: 3 ※Has FRQs?: Some Aero is a mild adolescent sounding voice with lots of noise. He generally sounds pretty sweet and light, I love his voice a lot! There is a decent anount of background noise, which some may be picky about. Still, it doesn’t affect the quality much and he works with most resamplers. Putting high BRE on him makes him sound pretty good. I generally had no issues with his oto, but the CC transitions aren’t otoed very well and some are off time.

Aino Erufu by Jasputing

※Reclist: 0.2.0  ※Number of pitches: 1 ※Has fully generated FRQs?: Yes, also has full llsm and pmk files Aino has a very unique voice acted timbre, and it’s almost difficult to find a song to match her. But when you do, she’s a sweetie. ^^ She has a sweet, louder voice and sounds best above A3, or even higher in the soprano range.  She is touchy with resamplers but works well with fresamp and world4utau.

Aki Toka/幸兔雪  by Yukito Yuki 

※Reclist: 0.2.0 ※Number of pitches: 1 ※Has fully generated FRQs?: Yes Aki is clear and strong and has a good tone. good tone, clear. has full frqs. got one of my favorite accents. Despite being monopitch, he has a very wide range and can go a little below his stated range of F3-C5. The mid/low range sounds best though imo. He comes with a guide for aliasing for those new to Arpasing, and a list of all the sounds which were affected by the VP’s accent. The Finnish accent does come through in the vowels and ending consonants, but the duplicates allow for more varied pronunciation so it’s not a big deal. Akihana ENG [radiance] D3_beta by Lunamageice ※Reclist: 0.1.0, uses phonetic filenames  ※Number of pitches: 1 ※Has fully generated FRQs?: Yes Akihana has a low, rich, full voice that’s good in bass and baritone range. I’ve used his VCCV before, but the Arpasing actually has more natural pronunciation. It has been in beta all this time, but it’s still worth it to use because of the rarity of this voicetype in UTAU. Angus by ReticentResolve

※Reclist: 0.2.0 ※Number of pitches: 1 ※Has fully generated FRQs?: Yes    I was blown away by how good Angus was! He’s a deep, soft masculine voice ranged F2-G4 with a nasal quality. He comes with a tripitch VCV JP bank for those who’d want that too. The mic he was recorded on picked up a lot of background noise and makes him sound muffled, but he’s still a decent vb.

AWG YukiList verKLAD by Adlez27 ※Reclist: YukiList with modifications ※Number of pitches: 1 ※Has fully generated FRQs?: No This is an example voicebank for the Yukilist arpasing recording style created by YukitoYuki. It has 800+ samples and a lot more coverage than basic Arpasing. It’s a tenor vb with a mild tone. The alveolar tap (dx) samples are in a separate folder for some reason and aren’t fully otoed, but they work fine. Some of the VC transitions aren’t configured correctly, but generally the oto is good.

Canary by SlightlyShredded

※Reclist: 0.1.0 ※Number of pitches: 1 (D3) ※Has fully generated FRQs?: Yes + llsm Canary has a really pleasant voice! He’s a sweet, fruity baritone with power in his voice. He’s very suitable for pop/edm. His oto kept spitting out errors while I used him but he sounds mostly fine despite that. 

Eon_Arpasing by fmccaddy12 ※Reclist: Custom (made by fmccaddy12) ※Number of pitches: 2 ※Has fully generated FRQs?: Yes + llsm Eon is a very realistic baritone voice. He sounds a little congested but is overall one of the best voicebanks on this list. He can be pushed a bit out of his range As far as I’m aware, his official site/character revamp is not done yet but his dl is still up on this website. Not even sure if I should share the link but it is still up publicly...

Hailey Denali by CatMoun

※Reclist: Custom  ※Number of pitches: 1 ※Has fully generated FRQs?: Yes + llsm  Hailey has one of the cutest tones one of all these voicebanks. She’s very nasal and sounds almost voice acted. Her reclist is 500+ samples long and doesn’t lack anything. However, all the vowels and most of the consonants are 2-3 syllables. Hailey also comes with an automatic updater.

鳳鐘ユウリ(Houshou Yuuri) by さきた (Sakita) 

※Reclist: 0.2.0 ※Number of pitches: 1 ※Has fully generated FRQs?: Some  Yuuri has a range of about A3-D5. She has a heavy Japanese accent but a decent range. She’s very emotional and pretty sounding, especially in the higher ranges. Though, if you want to use her, know that her oto requires a lot of fixing, as a lot of VC transitions are configured incorrectly.

Johatsun Kikuro Arpasing by Awes0meBr0/Biohazard-P

※Reclist: 0.2.0 ※Number of pitches: 1 ※Has fully generated FRQs?: Yes + llsm/bak/pmk   He has very quiet samples which causes lots of distortion in the program. Nevertheless, he still manages to sound realistic somehow-? He reminds me of older vocal synths like Microsoft Sam almost. I did mess around a bit with resamplers/flags for Kikuro and the best combination I’ve found so far was f2resamp and F0g1H20h0C10. Using Kikuro requires some effort but he’s sill pretty good.

Kara Kokone by SunGuardian524

※Reclist: 0.1.0 ※Number of pitches: 5 ※Has fully generated FRQs?: Some  Kara is most impressive for being one of the UTAU with the highest number of pitches on this list- each of them has a different tone, with one being a falsetto and another being a shout pitch. Her range is pretty wide, reaching almost into alto and being able to hit the high soprano notes too. However, this bank really does feel like a beta that hasn’t been touched in a while. It doesn’t have a fully configured prefix.map and the oto is pretty off and requires a lot of adjusting. I also wish she used a more comprehensive form of the reclist, since there is a lot missing. At least there’s a lot of duplicate entries.

Kayaka Konagi Arpasing(English) by こむぎこ (Komugiko)  ※Reclist: 0.2.0 ※Number of pitches: 1 ※Has fully generated FRQs?: Yes Konagi is very good, her pronunciation is clear and she has a nice tone. She is a bit finicky with resamplers and flags, probably meant to be used with default settings. She does have a Japanese accent but it’s a little less pronounced than Yuuri’s.

Kevin Futarine STUDIO (VP2 Arpasing) by FRAloids

※Reclist: 0.2.0 ※Number of pitches: 1 ※Has fully generated FRQs?: No  Kevin is another baritone voice with a youthful flavor. It has a good amount of flexibility, being sort of emotional and soft sounding but also capable of producing a bright, clear sound. He was recorded at E3 but has a flexible range and even sounds good higher. His pronunciation is pretty on point, but the t/d are a little mixed up. His dh sounds are also so short they’re hard to hear at times. The oto is generally good but might need some adjusting out of the box. He has a ton of extra samples of sighs, laughs, whispers, etc. I’ve been a Kevin fan for years and this voicebank definitely keeps the tone his Japanese vbs have.

Kii Nane by kii ※Reclist: 0.1.0 ※Number of pitches: 1 ※Has fully generated FRQs?: Some   Kii Nane is a sugary sweet voice that sounds best at C4 and above. Her samples have a lot of reverb, and there is some background noise which affects the output. Her pronunciation is very good, but a little slurred and off because of the engine noise. Resamplers are hit or miss, she does sound good with world4utau though.

Koyanagi Sayuki DEITY by Mango-tama

※Reclist: 0.1.0 ※Number of pitches: 4 ※Has fully generated FRQs?: No, but has full llsm  Sayuki is a very realistic voice ranged F#3-G5. Her low range is soft and her mid/high range is loud and clear. Lovely to tune and use in anything, I don’t have much negative to say about her.

Koyanagi Takehiko_TSUNAMI (BETA) by Mango-tama

※Reclist: 0.1.0 ※Number of pitches: 1 (G3) ※Has fully generated FRQs?: No, but has full llsm Takehiko is very cool sounding and bassy, and the voice acting makes him distinct from his sister. His best range is from D3 to G4 and he has sorta a soft tone that can be boosted via resamplers. He sounds great with f2resamp.

KYE by WinterdrivE

※Reclist: 0.1.0 + a few Alveolar tap samples ※Number of pitches: 5 ※Has fully generated FRQs?: No, but has full llsm Kye needs no introduction-he’s the guy who got most of us interested in Arpasing in the first place! His abundance of pitches works in his favor. He’s probably the best of these voicebanks to just plug and play to learn how to use Arpasing-if you don’t mind waiting for all those 1000+ frqs to generate. He is by far the best in baritone/bass range. A lot of the D4 pitch actually has pitch/frequency issues that can’t be fixed without workarounds, so using him from B2-C4 is your best bet-or just using a different pitch for the higher ranges. He does have a sort-of lisp but overall, not many oto or pronunciation issues.

Marie Kagekine Arpasing (v1.0) by AdagioFantasque

※Reclist: 0.1.0 ※Number of pitches: 1 ※Has fully generated FRQs?: Yes +  some llsm, pmk, and uspec  Marie is an operatic voicebank who tries to recapture the charm of Prima in an UTAU format. She has a wide range of B3-F5 and a mature, yet high and bright, sound. She has some of the best understandability out of any of these voicebanks. I really like her <3

Mikone Samu by codyboyo

※Reclist: 0.2.0 ※Number of pitches: 1 ※Has fully generated FRQs?: Yes Samu/Sam is a port of Microsoft Sam done in a similar way to how Defoko was made. He’s a crisp sounding masculine low voice. Due to the samples being short, he can sound very stilted. But for a jinriki (?) like this, he is very high quality.

Resonance Harmony (Arpasing English) by Hazu/パワ 

※Reclist: 0.1.0 ※Number of pitches: 6 ※Has fully generated FRQs?: Yes I have a lot to say about Harmony, and no good way of wording it. Okay. Resonance is very nasally and high, she has 3 base pitches plus growl, power, and whisper appends, and some vocal fry. Conceptually, she seems like a pretty good bank but I’ve always had trouble using her...it feels like a hit or miss on whether she’ll sound good or not. She sounds decent with f2resamp and w4utau but not many flags make her sound better. If anyone else wants to download her and play around, I’d love to hear it.

Rose Mae Arpasing BETA by Monochrome Rose ※Reclist: 0.1.0 ※Number of pitches: 1 ※Has fully generated FRQs?: Yes + llsm Rose has a bright, tomboyish, androgynous voice. The readme file notifies us of the non-native accent in the vb. It does have a drawl and some of the consonants aren’t recorded/otoed right, but overall the vb sounds very smooth. Honestly, I care more about recording conditions/oto/general feel than accent most of the time, and Rose does well in those ways.

Ryuunosuke Aoyama by Mattsuran 

※Reclist: 0.1.0 ※Number of pitches: 1 (B3) ※Has fully generated FRQs?: Yes + llsm Ryuunosuke is a childish, boyish voice with a strong Chilean accent. I think this vb may have actually been recorded by a kid, so that’s a point for uniqueness at least. He sounds best with fresamp and f2resamp. He has a gentle tone that can work in a lot of different ranges.

Tarek by Tadashi

※Reclist: 0.1.0 ※Number of pitches: 1 (B2) ※Has fully generated FRQs?: No, has llsm Tarek has a clear, loud tone with a noticeable lisp. UTAU recommends using him as a baritone, but I found putting him into higher ranges generally worked too. His samples sound less like singing and more like talking. He was recorded on a midquality mic which causes some distortion in the output. Still, I think his voice is really cute and I’m a sucker for baritones.

Thoma by Nythoma

※Reclist: 0.2.0 + vocal fry/alveolar tap ※Number of pitches: 1 (F#3) ※Has fully generated FRQs?: Yes   Thoma is an androgynous alto/mezo soprano. They have a soft, plasticy sort of feeling to their voice. They sound best from C3-F5. I feel a little..weird linking to/talking about Thoma since they were redesigned and are supposed to be a private voicebank, but the download is still linked on this site. Also please check out Nythoma when you get a chance, they make the best Arpasing USTs which helped me a lot while making this post. I feel like I haven’t checked in on anyone here in a while. How are you doing? I’ve written most of this post within the span of a few days and I think my brain is not capable of producing any thoughts other than “monopitch.”

Tsubasa Arugo Arpasing (BETA) by WinterStar

※Reclist: 0.2.0 ※Number of pitches: 1 ※Has fully generated FRQs?: Yes   Tsubasa has a soft, androgynous voice that sounds best from E3-A4. The samples are full of static, which causes some distortion. The ending vowel sounds also cut off too much at the beginning, making them too breathy. Still, it’s always nice to see another soft vb.

Xiao Meihua by 実偽Migi

※Reclist: 0.2.0 ※Number of pitches: 2 ※Has fully generated FRQs?: No Meihua has a realistic mature voice that’s also sweet and friendly sounding. She has a sort of low voice, and sounds the best from G3-C5. The download comes with a natural vb and a Sweet append, both monopitch. I listed them as being different pitches because they aren’t recorded at the same note. The Sweet voicebank is nasal and stronger too-it sounds more solid than sweet. She’s one of the vbs on this list that sounds the best with little to no edits. Her oto is smooth as butter :0

And that’s all of them done! Thank you SO much for reading if you got this far, seriously. You are what makes writing this blog worth it <3

Also I wanted to list my favorites of the bunch! This isn’t meant to pit any vb against each other or anything, just the ones I’m likely to go back to. -Canary -Eon -Xiao Meihua -Marie Kagekine -Kii Nane -Aki Toka -Thoma -Kevin Futarine Studio

Who are your favorites?