From the website: Chis-A can now sing in English! Provided as a default singer in VoiSona, Chis-A now has an English song voice library (paid subscription). Chis-A has an alluring, androgynous voice.
*cries in empty wallet*
*cries in empty wallet*
I wonder what it was about the English language that made them decide to hard pivot to a BDSM-lite bodycon look LOL. Reputation era Chis-A?! Very excited to hear her voice.
I was kinda hoping against hope Chis-A English might be free, but I guess the reasoning is that since you can input English lyrics into Voisona and have them converted into Japanese phonemes for her JP bank to sing, you're basically paying for the convenience of having a more natural pronunciation off the bat. (Part of me does feel like it would be nice if they released *a* free English bank at some point if they were interested in gaining more of an English-speaking audience, but I'm definitely not expecting them to do it.)
Either way I'm still hyped to hear her and maybe take advantage of the free trial campaign
Either way I'm still hyped to hear her and maybe take advantage of the free trial campaign
Can't wait to hear her bank, thoughj i guess it'll already sound fantastic considering her voicetype and her very own timbre
I don't think this product will be massively successful given the subscription model, but I still hope she sees some success and great songs. I think this will be possible thanks to VoiSona's increased focus on English localization compared to CeVIO CS and AI. I was also hoping she would be free-- that would really create the potential for English usage of VoiSona to really take off in the right environment. A fully-featured AI VB isn't even something Dreamtonics provides for free, so it would've been a major differentiator. Not surprised they didn't go that way at all, but it would've been cool. I expect they still probably see the primary userbase for this VB to be Japanese users.
Still, really excited to see how she sounds, I love the feeling of her Japanese vb and I'm really holding out hope that English will be great.
wrt her design, I think it's great, but I can't help but keep thinking...
(this just makes me like it more :) )
Still, really excited to see how she sounds, I love the feeling of her Japanese vb and I'm really holding out hope that English will be great.
wrt her design, I think it's great, but I can't help but keep thinking...
(this just makes me like it more :) )
I think I’d be a lot more adverse to the subscription model if it were only on a yearly plan. The monthly plan is only like $6/month and you can cancel it any time; that gives you a cheap and easy way to test her out, and if you don’t like her, you’re only out $6. If you DO like her, you’ve got the option of a yearly subscription or just paying a monthly fee when you plan to use her. For people who would use her heavily at first and then more sporadically over time (which is most people, if we’re being honest), it’s financially a better deal. It’s not a free trial, but it’s the next best thing: you don’t have to worry about finding someone to make renders for you or, worse, dropping $80-100 and realizing you won’t get much use out of the bank.
I get and sympathize with the objections to the model, but I genuinely don’t think it’s nearly as bad as people make it out to be.
I get and sympathize with the objections to the model, but I genuinely don’t think it’s nearly as bad as people make it out to be.
That's a fair viewpoint. It's a worse deal for power users of the VB, but a better deal for people that like to experiment with things. It's helps that the engine itself is "free".
I also feel similarly as in:
I don't think anyone is being unreasonable by not preferring the subscription based model, I even agree, but also I think that if parts of the community stopped for a second, many would realize subscription based model isn't inherently incompatible with their already existing habits, as in: it's extremely common for people in the community to purchase a vb and use it for few months only to resell it because they no longer have use for it after few usages.
Logically, it shouldn't be that big of an issue, but it's that lifetime purchase gives greater sense of control, and subscription based model becomes an issue when you come to love the vb and want it to stay for prolonged time, or as Pico said: "power users of the VB", so I share the sentiment that it's not my preference, but also as you said, it doesn't pose as many issues as it's made out to be, at least to great majority of western userbase, but I don't expect the feelings on subscription based models to change despite that.
I got jumpscared by the orange thing. But yes, I do agree that by making chisa eng paid, ts really is missing out in an opportunity to attract a whole new audience to Cevio and voisonaI don't think this product will be massively successful given the subscription model, but I still hope she sees some success and great songs. I think this will be possible thanks to VoiSona's increased focus on English localization compared to CeVIO CS and AI. I was also hoping she would be free-- that would really create the potential for English usage of VoiSona to really take off in the right environment. A fully-featured AI VB isn't even something Dreamtonics provides for free, so it would've been a major differentiator. Not surprised they didn't go that way at all, but it would've been cool. I expect they still probably see the primary userbase for this VB to be Japanese users.
Still, really excited to see how she sounds, I love the feeling of her Japanese vb and I'm really holding out hope that English will be great.
wrt her design, I think it's great, but I can't help but keep thinking...
(this just makes me like it more :) )
on a lighter note, I really like the text at the bottom
also
also
I really wish/hope Cevio will implement actual honest to god cross lingual synthesis (not that glorified phonemizer they call cross lingual, yes I am still salty) cause I really like a lot of cevio voices, I just wish I could use them in English (give me Yume no Keshou Rose with english cross lingual please please please). Or at least the more popular banks getting english vbs.
Honestly, without sufficient English-language data, a crosslingual system wouldn’t be much different from the current system they have in place. The trick would be finding a VP who could provide English data, or getting current VPs to learn sufficient English, otherwise it would just be redundant. I’d like to see future implementation of such data! But what we have now is the closest equivalent without telling current VPs “Hey, we’ve got a few fans wanting to use you for English, can you go out and get an A1 certification in English for us?”, something that’s not really sustainable on such a heavily Japanese-marketed synth.
TechnoSpeech and Kamitsubaki cevio ai vbs are now on Sonicwire
sonicwire.com
「CeVIOプロジェクト」「KAMITSUBAKI STUDIO」製品の取扱いを開始しました！ | SONICWIRE BLOG
この度SONICWIREでは、「CeVIO AI（チェビオ エーアイ）」を使用した製品を多数展開する「CeVI
Chis-A English Demo!!
So Personal opinion. This will change after listening to the next demos, and the upcoming song by CircusP, an actual english speaker who have experience tuning English songs.
WAY better english pronunciation as expected... BUT there is a slight accent on specific consonants. But it's like 5% there and on the high notes, too small to be noticed but as someone who tried to make JP Chis-a sing english, it bugged me. Still, it is a huge improvement and adds some character to her.
She has that androgenous and lower timbre just like her Japanese voice, which is good and very consistent. Except when she tried to hit the high notes but it's just a tiny nitpick. Though I'm also curious to know just how much editing was done. Was this 100% AI or there was like, 20% manual editing? Things can change in the hands of someone else.
So Personal opinion. This will change after listening to the next demos, and the upcoming song by CircusP, an actual english speaker who have experience tuning English songs.
WAY better english pronunciation as expected... BUT there is a slight accent on specific consonants. But it's like 5% there and on the high notes, too small to be noticed but as someone who tried to make JP Chis-a sing english, it bugged me. Still, it is a huge improvement and adds some character to her.
She has that androgenous and lower timbre just like her Japanese voice, which is good and very consistent. Except when she tried to hit the high notes but it's just a tiny nitpick. Though I'm also curious to know just how much editing was done. Was this 100% AI or there was like, 20% manual editing? Things can change in the hands of someone else.
Last edited:
The accent doesn’t strike me as a flaw. If I wanted a voice that sings English with an American accent, I’d use an American English VB. 
I think she sounds astonishing! Aaaaaah I love her so much~
I think she sounds astonishing! Aaaaaah I love her so much~
She sounds so nice! I would be open to doing a sub thing for a while when I got the hankering to use her, but I do still wish there were a way to just one-time purchase her, even if it was expensive. It's just a weird personal pet peeve, but with a sub-only style, I would never feel like I truly "had" the bank, if that makes sense.
Second English Demo for Chis-a!!
This is the english version of her previous song Heart flower or ちいたな
Now I'm excited to hear an OG song demo.
This is the english version of her previous song Heart flower or ちいたな
Now I'm excited to hear an OG song demo.
Here is a Voisona tip video I found in youtube. Unfortunately, it is completely in Japanese but there might be some tips to that can be reviewed over to help in using Chis-a better. If there are any translators, your help is very much welcomed.