SynthV SynthesizerV Musical Isotope KAFU

[nightmare]

Hopefully, it doesn't end up with SynthV becoming the monopoly. Some SynthV fan are really pissing me off with their constant whining that X Character is not on SynthV and now they're are overjoyed to see more companies potentially abandoning y synth program for their choice.

Vocaloid has had a monopoly on vocal synths for an extremely long time. An engine that produces clear vocals and is easy to use should gain popularity. SynthV AI is newer than Vocaloid or CeVIO, so it's getting voicebanks for popular mascots last.

(Credits to @LilyAnne (EDIT 3: I can't seem to find their VocaVerse profile. They are on Discord, though) for discovering the demo before me. I was originally going to post the demo here, but I was too late, it seems.)

Huh. SynthV KAFU sounds like a mix between KAF and CeVIO AI KAFU (unsurprisingly) to me. My only issue with CeVIO KAFU was that she sounded way too nasally and congested for my ears. SynthV KAFU at least resolves that. Although, if you're a fan of CeVIO KAFU's tone/timbre, this probably won't be for you.

Sorry, I forgot my nickname is different here haha.

I hated how congested and nasal CeVIO Kafu sounded. CeVIO's engine also makes voices extremely fuzzy, equivalent to seeing someone behind a fog. SynthV Kafu sounds lovely in comparison.

 

[Blue Of Mind]

Does Kafu SynthV have a soft or whisper vocal mode, or is that how her default is meant to sound?

Aside from that, she does sound very different from her CeVIO - probably the most 'unlike' port to SynthV we've had yet, and I remember seeing a bunch of complaints about Teto SynthV despite her tone not sounding that different compared to her UTAU.

 

[nene]

She sounds a bit breathy here? Never listened to the original, though. Hope more demos come out, maybe even one in a different language.

I feel like the tuning for phony has always been kind of breathy, but maybe we'll get a raw demos/new songs later?

(Or at least something like a re-cover of Cute no Kanojo that's very bright & clear to match.. )
~~~~~~~~~~~~~

Phony has always been a favorite song to me so it's crazy how close it sounds minus the engine noise??? Clear too! i'm in love!

 

[Mika]

She sounds amazing! Definitely seems like they're using multiple vocal modes in the song. She lost a little bit of her iconic sound, but she feels more different from Rim now, at least in my opinion? Their cevio banks sound so close, just with different pitches, this feels different. And I think with the right settings and tuning you can get her to sound more like her cevio voicebank.

 

[hierco]

I'm absolutely in love, I've never been a huge fan of that twang she has on CeVIO (no disrespect for the program itself), but Kaf has such a pretty voice so to hear a more 'natural' version of KAFU makes me so happy. Can't wait until she comes out!! I agree with her sounding quite breathy, which makes me really curious about her vocal modes...

 

[Maxi]

I really like how she sounds !

Her Amazon listing says her vocal modes will be: Emotional／Kawaii／Stylish／Powerful
I wonder if "Emotional" is the breathy sounding one we listened to in the demo

Amazon listing

 

[lIlI]

If they're aiming this update at me specifically, they did a great job, as they removed the precise elements I didn't like about her CeVIO, haha. She no longer has that nasal froggyness that made her previous voicebank so divisive with Western fans. However, I think if they're smart their next demo should use tension/GEN/her Kawaii mode, to show that she can get closer to her original tone. While I didn't like it, it was undeniably a huge hit with her main audience.

SynthV is an extremely flexible engine, but sometimes companies do a truly terrible job of showcasing its range. I don't doubt you can modify Kafu to sound more like her CeVIO, but her company needs to showcase this for people not familiar with the engine. Dreamtonics is without a doubt the worst offender- we just had Ayame, who has an idol vocal mode that sounds completely different to her base tone, giving her a cutesy timbre that's really popular in Japan...and wasn't shown at all in marketing. Fingers crossed Kamitsubaki has lots more demos to come.

I'm also very interested to hear Kafu's English, Kaf's singing style is so unique. The closest English equivalent I can think of is Ellie Goulding, it would be fascinating to get something similar.

 

[blackout]

LOVE her, this first sample did not disappoint at all despite the long wait. She’s definitely closer to Kaf, thus more natural sounding than Cevio Kafu. She’s also different enough from her Cevio AI in a way that doesn’t really replace it, I think both versions of her will have their uses!

I cannot wait to hear more, I really hope we get to hear some cross lang samples as well!!

 

[pico]

Obsessed with how good this sounds in a way that is specifically not a stereotypical "Synthesizer V voicebank" way. My least favorite characteristics of SV as an engine are its piss-poor whisper renderings and its overeliance on the "base model" singing sound for consistently nice-sounding voices, rather than voices that retain the original singer's cadence. Either this demo was meticulously post-processed, or they managed to work with Dreamtonics to create something truly cool here.

I figured Dreamtonics would not allow the "formant shift" lock that they used on *CeVIO so I'm really glad the compromise was this "pitch snap" sound. It's adorable. This is what I want to see from AI voicebanks-- unique stylistic sounds that can only come from a computer, with the added bonus of quality engine rendering that creates the output the developers intended. THANK YOU Kamitsubaki.

 

[AmazingStrange39]

If they're aiming this update at me specifically, they did a great job, as they removed the precise elements I didn't like about her CeVIO, haha. She no longer has that nasal froggyness that made her previous voicebank so divisive with Western fans. However, I think if they're smart their next demo should use tension/GEN/her Kawaii mode, to show that she can get closer to her original tone. While I didn't like it, it was undeniably a huge hit with her main audience.

SynthV is an extremely flexible engine, but sometimes companies do a truly terrible job of showcasing its range. I don't doubt you can modify Kafu to sound more like her CeVIO, but her company needs to showcase this for people not familiar with the engine. Dreamtonics is without a doubt the worst offender- we just had Ayame, who has an idol vocal mode that sounds completely different to her base tone, giving her a cutesy timbre that's really popular in Japan...and wasn't shown at all in marketing. Fingers crossed Kamitsubaki has lots more demos to come.

I'm also very interested to hear Kafu's English, Kaf's singing style is so unique. The closest English equivalent I can think of is Ellie Goulding, it would be fascinating to get something similar.

ooh, i've always loved ellie goulding's voice so an ellie-goulding sounding vsynth would be amazing

 

[Leon]

As expected, I've seen some argumentation on twitter about whether she's higher quality or lost some of her charm. I guess as someone who didn't ever care for her CeVIO bank (I'm sorry, how she -- the voicebank, not the human singer -- got SO popular is just beyond me), I would say both are right and that's a good thing? But I also think that it's too early to judge from one demo with autotune applied...

 

[pico]

I really think the "autotune" sound isn't post-processing but instead an affect of the VB intended to be like what we heard from the "third mode" proposed for CeVIO AI during the voting period. KAF insisted that KAFU be distinct from her voice, so I believe this is the compromise that was made. Only time will tell though.

 

[Leon]

That's possible I guess? But I would really think (slash, hope?) that they would've informed people before preorders if the bank was permanently in autotune mode. I really can't see DT approving that without making it expressly clear that that was the gimmick. I'd like to see another demo before we jump to any solid conclusions, anyway :)

 

[pico]

In fairness, KAFU has always been a product “with a gimmick”, from the moment she was announced to the present day, so I think it’s reasonable for consumers that pre-ordered without ever having heard a demo to at least have held onto the context that the product has always been marketed as KAMITSUBAKI’s interpretation of an “isotope”— an artificialized “snapshot” of someone’s voice.

But I agree, excited to hear more demos!

 

[Leon]

I guess you're right. I was thinking from my own POV, if I ordered her thinking it would be something like the CeVIO bank but on an editor I already own, or something like that, I'd be pretty mad if it turned out that she was always on autotune without prior warning. But regardless, she did sound good, so I hope KAFU fans will make good use of her when she drops!

 

[nene]

Its been reported that people who preordered KAFU SV, now formally titled The Right Stuff Ver., are recieving emails saying she's going to be delayed (unsurprisingly).



As of now, she has no release date.

 

[Blue Of Mind]

<snip>

I've been kind of expecting this, considering her first demo appeared close to the original release date. But the lack of a new date is really making me wonder what's been going on behind the scenes - I don't think we've had a SynthV release this troubled and lacking transparency before.

 

[mary34]

Well I mean there was the boxed version of Eleanor Forte AI where customers basically had to ratio VOLOR on social media into announcing the boxes were canceled after several months of radio silence. I would say this is comparatively not as bad.
I am curious what the reason could be thought. Doesn't seem to be a simple case of a delay due to needing more time to finish because 1. they would have said so and 2. they would have given a date (or at least a rough estimate of when to expect it). The fact that they are offering preorder refunds while neat also has some troubling implications: I am not familiar with the Musical Isotopes and how Kamitsubaki does things in general but this offer signals to me they aren't even sure if they can release the product due to whatever is causing the delay (or at least not in the near future). This is pure speculation and I don't want to cause alarm but I'm a little worried

 

[Prism]

This is 100% speculation

I hope the delay isn't because kafu sounds too close to kaf and this is a consent problem.

 

[junky]

This is 100% speculation

I hope the delay isn't because kafu sounds too close to kaf and this is a consent problem.

This is what I suspected too.