Holy heck, I looked at other videos on that channel and I hate them all. Somehow the unsettlingly ye old 3D-looking characters dancing to music with bland voices and crazy fast BGM annoys me, in addition to the fact that their mouths don't move correctly.
The thing this video is from is called COPIPEandROID (kopipe = copy/paste, & = and, roid = roid, so it's "Copy-Paste Android"). The app lets you customize a character to dance in 3D and you basically use photos you own to make a slideshow. It's aimed at high school girls who want to say what's truly in their heart, but can't. This is a commercial explaining how to use the app (2nd video, not the 1st):
Anyway, I did find out it is using CS7 and Satou Sasara in this official tweet thread by Techno-Speech:
The thing this video is from is called COPIPEandROID (kopipe = copy/paste, & = and, roid = roid, so it's "Copy-Paste Android"). The app lets you customize a character to dance in 3D and you basically use photos you own to make a slideshow. It's aimed at high school girls who want to say what's truly in their heart, but can't. This is a commercial explaining how to use the app (2nd video, not the 1st):
Anyway, I did find out it is using CS7 and Satou Sasara in this official tweet thread by Techno-Speech:
Interface revealed for CeVIO AI, announcement coming soon from Techno-Speech.
The UI looks...well, to be completely honest, messy and confusing. I do not know if it is all the different colors or a problem with the UI itself, but it is difficult to look at personally.
It is nice that they are trying out a different style, but I think I prefer the older CeVIO UI. Even if it did have that same rainbow keyboard set up that also is hard on the eyes.
I wonder who that silhouetted figure is. The hairstyle looks like Sasara's, but we have not heard anything about her coming to AI yet. And why would they try to hide that it is Sasara by giving her a silhouette portrait if it is her?
It is nice that they are trying out a different style, but I think I prefer the older CeVIO UI. Even if it did have that same rainbow keyboard set up that also is hard on the eyes.
I wonder who that silhouetted figure is. The hairstyle looks like Sasara's, but we have not heard anything about her coming to AI yet. And why would they try to hide that it is Sasara by giving her a silhouette portrait if it is her?
I assume you will be able to change the eye-burning rainbow to their more neutral greys like current CeVIO. It's really disorienting for me to try to use the rainbow one.
Yeah, I thought it reminded me of Sasara, too. Having it be a silhouette reminds me of what Crypton sometimes does with their icons. I guess they want to build hype and not spoil stuff.
Yeah, I thought it reminded me of Sasara, too. Having it be a silhouette reminds me of what Crypton sometimes does with their icons. I guess they want to build hype and not spoil stuff.
I did not know CeVIO had the option to change colors. Hopefully they will carry that option into AI and Pro. I have not gotten to use CeVIO yet, but the rainbow makes me dizzy just seeing it.
I cannot imagine how it would be to work on a project with that.I forgot Crypton used to do that. Perhaps Sasara will be getting a new design then? Crypton's silhouette's used to be for new designs, so I would assume the same is happening with Sasara. It would not make a lot of sense to make it a mystery and it turns out that it is just the same old Sasara
I still find it odd that she was not part of the initial announcement lineup. Perhaps she will come with the engine as a default vocal?
They uploaded a small demo on twitter, I'd call this a teaser since it's like 5 seconds
That voice is definitely Sasara
That voice is definitely Sasara
This confirms Sasara along with customizable UI(notice that the rainbow is missing). The UI is still a little rough, but it is looking better. Hopefully it will have more customization options. The random purple in the UI looks odd.
omfg I hate to be posting in so many different threads at once, but here's an example of AI Talk that AHS uploaded.
(Listen to the end, there are English voices as well.)
(Listen to the end, there are English voices as well.)
The English sounds about like Microsoft voices or Google translate. Actually Microsoft Cortana sounds clearer but then the sentences are pre-recorded.
I'm surprised there's a british accent setting, still autotuned but sounds like Oliver's aunt speaking. Even the noise makes it sound like it's was made same year as when Oliver was released, and that was ages ago.
I'm seeing the possibility of IA English talk with a British accent.
I'm surprised there's a british accent setting, still autotuned but sounds like Oliver's aunt speaking. Even the noise makes it sound like it's was made same year as when Oliver was released, and that was ages ago.
I'm seeing the possibility of IA English talk with a British accent.
Last edited:
This is kind of over the top, but even if the TTS English voices sound normal/slightly boring, I would definitely buy one if one becomes available, because having to switch to websites to try to download English TTS clips and go back to VOICEROID/CeVIO is really annoying. It would be nice to have JP and ENG TTS in ONE INTERFACE. (I know this problem is highly specific.)
Edit: Oh yeah, it says on the video that the talk voice is a prototype and they intend to raise the quality.
Edit: Oh yeah, it says on the video that the talk voice is a prototype and they intend to raise the quality.
I could be very wrong, but there's a chance the English vocals demonstrated were sampled from the Japanese voicebank and use CEVIO's ai to replicate the diction and pronunciations of an English speaker. That would account for the quality drop off during the language shift at least. I could be very wrong though.
I don't think that the English vocals are edited Japanese ones, because they would have a noticeable accent (Japanese lacks many vowel phonemes we have, they only have 5 vowels and that would be very hard to edit) and it would be jittery. I think it's more likely they downloaded English voice data from a corpus that offers free files for deep learning purposes or something.
It bugs me a bit on why AHS is releasing a longer Cevio AI demo video.
Cevio released a 5 second video in twitter but AHS released a longer demo of the same tech but it shows 4 different voice settings including a beta English prototype (that will probably be commercially ready in 2 years if you consider how long it took for Cevio to release any news on the early AI tech demo last year)
I'm guessing they have a some usage deal, but it shouldn't be a surprise since both the companies specialty complement each other.
AHS will be using the AI tech in their products so is 1st Place. Cevio also uses IA in their websites for sample demonstration.
But judging with how AHS included the English talk prototype, they are possibly planning to make an English AHS talk bank character in the future, with a singing bank too.
Cevio released a 5 second video in twitter but AHS released a longer demo of the same tech but it shows 4 different voice settings including a beta English prototype (that will probably be commercially ready in 2 years if you consider how long it took for Cevio to release any news on the early AI tech demo last year)
I'm guessing they have a some usage deal, but it shouldn't be a surprise since both the companies specialty complement each other.
AHS will be using the AI tech in their products so is 1st Place. Cevio also uses IA in their websites for sample demonstration.
But judging with how AHS included the English talk prototype, they are possibly planning to make an English AHS talk bank character in the future, with a singing bank too.
Do you think it's possible that Zunko could end up being this character? I mean, we know she will be getting an English talk bank at some point so this would be the perfect opportunity for them. This is just pure speculation, though.
We already know AHS has been planning English talk voices. Zunko was the first confirmed one, but they said they had plans for others as well.
Only thing is... all of those were announced for VOICEROID. It's really weird to me that they'd abandon their own product line for pretty much their only competitor's, but they were the ones who encouraged Rikka to move to CeVIO in the first place, so... I have no idea what's going on.
(In Zunko's case though, I'm sure she'd be thrilled to be on both VOICEROID and CeVIO whenever they can get her English VP's situation sorted out, so I don't think there's much to worry about regarding her.)
Only thing is... all of those were announced for VOICEROID. It's really weird to me that they'd abandon their own product line for pretty much their only competitor's, but they were the ones who encouraged Rikka to move to CeVIO in the first place, so... I have no idea what's going on.
(In Zunko's case though, I'm sure she'd be thrilled to be on both VOICEROID and CeVIO whenever they can get her English VP's situation sorted out, so I don't think there's much to worry about regarding her.)
Today I saw a tweet where AHS announced that CeVIO will soon be compatible with their video editing software, Recotte Studio (which is directly linked to VOICEROID for making audio with 2D and 3D talk/Let's Play videos). Do you think that maybe AHS is turning into some sort of... connector between all of these other companies? Synth V, CeVIO AI, VOICEROID/AITalk. Seems like something weird is going on.
I legitimately don't know which thread to post this in. I haven't been super active so I'm not sure if someone noticed this already, but I gathered some information that makes me feel worried about flower for CeVIO AI.
The Gynoid Twitter person said they had worked for Gynoid for 4 years, and Xin Hua came out in 2017 (isn't that like 4 years, too?). I am hoping there are more than one employee at Gynoid... S:
The Gynoid Twitter person said they had worked for Gynoid for 4 years, and Xin Hua came out in 2017 (isn't that like 4 years, too?). I am hoping there are more than one employee at Gynoid... S: