I came here because of IA. Because IA is my favorite Vocaloid. I admit that CeVio's Voicebanks sounds very natural and close to real singers. But Vocaloid in particular is mire artistic, specifically because of its unauthenticity, yet high quality, which makes Vocaloid itself a special segment of musical art.
@@victorp8689 I think they so use being top that they felt never going challenger on their level especially in the English market. they mess up on v5 allowing synthv get an edge on the English market. I think only reason did not learn about synthv for awhile mine interest in sing synth waned now try get replacement for my old Vocaloid because mine old computer died.
@@victorp8689 I beg you to listen to good robotic tuning and learn about opinions, realism doesn't have to be the goal. If you want a real voice, go listen to an authentic singer.
3:39 Oh god, I’ve forgot that I’ve tweeted than in 2018. Now seeing it, I remember being very upset that IA migrated from VOCALOID to CeVIO at the time. To be really honest, the engine noice was very loud in early CeVIO and wasn’t huge fan of it. However, after CeVIO AI and KAFU release, the engine noise has improved significantly to the point that doesn’t bother me anymore. Heck, I’m even interested to get CeVIO AI if Mac OS ver was available. Not to mention that V5 has been decline of VOCALOID as software… so I think that IA moving to CeVIO might’ve been the smartest choice in long run. I didn’t realise it at the time I tweeted that, I was unnecessarily upset. Interesting to see own opinion change over time and great video btw. I truly got goosebumps seeing an old tweet of mine from an old account that I’ve since deactivated to be presented in someone’s video.
this production quality of this is so impressive, i knew it would be good but you really blew it out of the water!! wonderful work davee, im obsessed with this new style and tender care put into your videos. you've grown so much, keep shooting for the stars!!!!
I recently heard niki's new song UraSekai ft. Kafu. I thought Kafu might be another cover artist (like Utaite) but it was actually an AI and here I am learning about this CeVIO AI. AI covers been really cool these past few years and I love it. (Although, respect to the original artist) Kinda feels cool growing up from Vocaloid / Nico Nico era and to this stage.
As an Utsu-P fan (albeit a newer one) I actually really liked Atari Front Program and PEE PEE PRINCESS, both songs using KAFU. The new voice was really refreshing and honestly was a really cool contrast to the messy instrumentals /p being let loose in the two songs. Frankly I'm excited to see how Utsu-P uses this voicebank further, or if he'll have her duet with some other vocaloids he used to use more often such as GUMI or Rin. Overall I'm not disappointed or worried about overeliance, I'm just excited as all hell for what's to come! :D
@@youngsoundpi kinda disagree cuz when i think of cevio i think of kafu rather than ia. when i think of ia i think of her as a vocaloid cuz that’s what she originally was.
Just a heads up, the CeVIO AI samples you used around 5:01, are a comparison between AI and CS7, and it's using a different vocoder than the release version, and up to now, so it sounds clearer than the actual CeVIO AI program. Overall, nice video!
i would greatly argue that CeVIO gained its prominence way back when big name producers like ATOLS & Police Piccadilly were using ONE. Not only that but CeVIO also gained alot of notoriety from its talking voices regularly used by Nico Nico users with gaming narration. Kafu came SUPER late to the party and was not the reason why ppl knew about CeVIO and used it. but i would say the bank encouraged ppl to get CeVIO AI specifically if not for Yukari Rei or IA English AI already (not to mention Satou AI and IA JPN AI)
It's been months of abstinence but he is finally back. Santa is real and his name his davee jonesey. Thank you for this gift of an informational video.
Your editing is so godlike, Davee. I can't wait to see what other content you'll come out with in the future. Honestly at a loss for words rn, so I just wish you the best. Please don't forget me when your channel blows up!!
you explained it so well, now I finally understands more about cevio. I wasn't aware of it before one of my fav producer, iyowa, uses it. And honestly almost didn't notice until I checked again the vocal then sees KAFU, since their tuning is usually unique from other vocaPs. I think it what makes kyukurarin popular as well, instead of just relying on AI in cevio, producers should tune using their style. I've seen some new cevioai users who didn't tweak much and makes their voice sounds average to others wwww
Thank you for educating me!! It's difficult to keep up with all the vocal synthesizer technology, and I had no idea about CeVio. But now I do, so thank you!!
Wow I’m doing research on Vocaloid and other Vsynth companies and this was SUPER helpful since I’ve been absolutely into Kafu as of late! I was even having the sort of same feeling about AI voices as of late, mostly with the release of CiFlower, but now that producers are showing what they have to offer maybe it won’t be so bad after all, plus I know that I’m at least not alone with what I thought, so keep up the good work 🎉🎉🎉
Hello hi, im grateful for this video since its edited well and its catchy But i want to put in some details about how CeVio works. It uses HMM like Vocaloid 1 but onwards vocaloid stopped using it. HMM observes some recordings and recreates them while vocaloid 2 onwards takes the recordings and stretches them like UTAU. Cevio with ONE added auto-tuning to non AI vbs and about the tuning this isn't like we expect, it just averages the pitch between silence and vowels. It's mostly extracted from modulation so you have to tune AI voices the same way like you do with HMM but the difference is that AI will re-render the note more naturallay this means it extracts data like strength-breath-accent based on the song data it has a similar pitch bend to what you did in theory. edit: I wanna add about backround noise. HMM produces more backround noise since it recreates a whole sound through software V1 and CevioCS have a similar air like engine noise. V2 onwards and UTAU produce backround noise is the sample needs to be stretched - vocaloid uses exact cuts and loops so noise woudnt be created while UTAU will produce a metallic sound usually the example you showed didn't have engine noise, the backround noise is actually the formants that are peaking and resonating in the recordings so that's a microphone problem, they are usually heard louder in utau than the recording cuz of compression.
Woah that's interesting. I apologise for the lack of a clear explanation and my mistakes in the video, I don't know much about this sort of stuff so this is quite the insight. Thank you for your comment!
Wow, this is amazing! The production quality is of course through the roof, and I learned lots of things I didn’t know about CeVIO before! Do you think you could do one on Synth V too? I think they get some really impressive results, especially with Solaria’s recent release. You got a new subscriber! Also one thing I’ve heard people complain about with CeVIO is that they can’t sing as fast as other synths, even other AI synths like Synth V AI. Past a certain speed their engine noise and failure to properly enunciate syllables makes them impossible to understand. Also this is my opinion, but I think they have a tendency to sing flat as well. This can be fixed of course, and for some voice banks it works in their favor (most notably KAFU) but these issues contributed, I think, to issues in other voicebanks, including IA AI’s lukewarm reception
8:45 this very much. i love how kafu sounds, but yeah it does start to sound very similar across all artists… basically kafu becoming more like a digital utaite rather than a voicebank. which isn’t a bad thing! but it’s different!
I like the fact that CeVIO AI and SynthV have a higher quality and realistic voices. However, I none of the songs makes with this voices got attached in my brain like vocaloid songs. Maybe it's because VOCALOID voices are more catchy and more addictive to listen to.
So I was wrong for 4 years... I thought that CeVIO is a part of vocaloids... Like the way of tuning or smth makes them special, not-like-the-other-vocaloids
Cevio sounds very realisticly when used right. But even producers like ATOLS can't get rid of the "nose tone". The singers sound like they have a stuffed nose and sing through it. This is why I am very puzzled if I rather have a clear robotic voice. Or a stuffed nose realistic voice.
Also! CeVIO is largely based on Sinsy, (and CeVIO *is* basically commercial Sinsy), so if you wanna see the development of CeVIO, you should look at Sinsy's first.
I was so excited when you said machine learning earlier in the video but then you switched to calling things AI which doesn't exist 😭 I understand calling it Cevio AI as that's the name of the program even if the last part of the name is untrue. AI would be things like HAL 9000 from 2001 a Space Odyssey, Cortana from Halo, or Glados along with the cores and turrets from Portal. The only reason companies call what we have now AI is because it rolls off the tongue better and is catchier but they're still calling a plant a human.
In my opinion I really dislike CeVIO…I do think it is a step down from before. Most voices like kafu and also hearing Flowers demo I really dislike them. And now with all these Producers switching over to CeVIO I’m starting to get a little borde and starting to really hate CeVIO more
I came here because of IA. Because IA is my favorite Vocaloid. I admit that CeVio's Voicebanks sounds very natural and close to real singers. But Vocaloid in particular is mire artistic, specifically because of its unauthenticity, yet high quality, which makes Vocaloid itself a special segment of musical art.
i beg you to listen to others voicesynths apart from vocaloid (like synthV), vocaloid Is so behind and IT SUCKS
@@victorp8689 I think they so use being top that they felt never going challenger on their level especially in the English market. they mess up on v5 allowing synthv get an edge on the English market. I think only reason did not learn about synthv for awhile mine interest in sing synth waned now try get replacement for my old Vocaloid because mine old computer died.
@@victorp8689 I don't think approaching total realism should be the goal of every single synth.
@@victorp8689 I beg you to listen to good robotic tuning and learn about opinions, realism doesn't have to be the goal. If you want a real voice, go listen to an authentic singer.
3:39 Oh god, I’ve forgot that I’ve tweeted than in 2018. Now seeing it, I remember being very upset that IA migrated from VOCALOID to CeVIO at the time.
To be really honest, the engine noice was very loud in early CeVIO and wasn’t huge fan of it. However, after CeVIO AI and KAFU release, the engine noise has improved significantly to the point that doesn’t bother me anymore. Heck, I’m even interested to get CeVIO AI if Mac OS ver was available.
Not to mention that V5 has been decline of VOCALOID as software… so I think that IA moving to CeVIO might’ve been the smartest choice in long run. I didn’t realise it at the time I tweeted that, I was unnecessarily upset. Interesting to see own opinion change over time and great video btw.
I truly got goosebumps seeing an old tweet of mine from an old account that I’ve since deactivated to be presented in someone’s video.
AND WHEN THE WORLD NEEDED HIM MOST HE RETURNED!
this production quality of this is so impressive, i knew it would be good but you really blew it out of the water!! wonderful work davee, im obsessed with this new style and tender care put into your videos. you've grown so much, keep shooting for the stars!!!!
I recently heard niki's new song UraSekai ft. Kafu. I thought Kafu might be another cover artist (like Utaite) but it was actually an AI and here I am learning about this CeVIO AI.
AI covers been really cool these past few years and I love it. (Although, respect to the original artist)
Kinda feels cool growing up from Vocaloid / Nico Nico era and to this stage.
As an Utsu-P fan (albeit a newer one) I actually really liked Atari Front Program and PEE PEE PRINCESS, both songs using KAFU. The new voice was really refreshing and honestly was a really cool contrast to the messy instrumentals /p being let loose in the two songs. Frankly I'm excited to see how Utsu-P uses this voicebank further, or if he'll have her duet with some other vocaloids he used to use more often such as GUMI or Rin.
Overall I'm not disappointed or worried about overeliance, I'm just excited as all hell for what's to come! :D
Kafu is Cevio's Hatsune Miku, basically
that would probably be ia, id say kafu is cevio's gumi
@@youngsoundpi kinda disagree cuz when i think of cevio i think of kafu rather than ia. when i think of ia i think of her as a vocaloid cuz that’s what she originally was.
Just a heads up, the CeVIO AI samples you used around 5:01, are a comparison between AI and CS7, and it's using a different vocoder than the release version, and up to now, so it sounds clearer than the actual CeVIO AI program. Overall, nice video!
Ah, I apologise for that mistake. I didn't realise the other one was CS7. Thank you for letting me know!
i would greatly argue that CeVIO gained its prominence way back when big name producers like ATOLS & Police Piccadilly were using ONE. Not only that but CeVIO also gained alot of notoriety from its talking voices regularly used by Nico Nico users with gaming narration. Kafu came SUPER late to the party and was not the reason why ppl knew about CeVIO and used it. but i would say the bank encouraged ppl to get CeVIO AI specifically if not for Yukari Rei or IA English AI already (not to mention Satou AI and IA JPN AI)
A nice informative video! I hope you do an "all about Synthesizer V" in the future, davee.
Maybe one day
It's been months of abstinence but he is finally back. Santa is real and his name his davee jonesey. Thank you for this gift of an informational video.
Your editing is so godlike, Davee. I can't wait to see what other content you'll come out with in the future. Honestly at a loss for words rn, so I just wish you the best. Please don't forget me when your channel blows up!!
I can't believe all this time i pronounced cevio wrong 😔😭
OKAY SO IT'S NOT SEEVIO?? 😭😭😭😭
very epic, liking the style experimentation alSO PIAAAAAAAANOOOO MAAAAAAANNNNNNNN
Awesome production value, I have known CeVio and their products for a while but this amount of information was extremely enlightening.
you explained it so well, now I finally understands more about cevio.
I wasn't aware of it before one of my fav producer, iyowa, uses it. And honestly almost didn't notice until I checked again the vocal then sees KAFU, since their tuning is usually unique from other vocaPs. I think it what makes kyukurarin popular as well, instead of just relying on AI in cevio, producers should tune using their style. I've seen some new cevioai users who didn't tweak much and makes their voice sounds average to others wwww
holy shit i just got goosebumps with AI tuning
Wouldn't have known anything about Cevio if davee didn't make a video on it.. Thanks man
Sometimes, Patience is Virtue. Amazing job Davee. Thank you for your OPEN-MINDEDNESS :)
Thank you for educating me!! It's difficult to keep up with all the vocal synthesizer technology, and I had no idea about CeVio. But now I do, so thank you!!
imagine they name the cevio ia ''ia ai'' that would be really funny
Wow I’m doing research on Vocaloid and other Vsynth companies and this was SUPER helpful since I’ve been absolutely into Kafu as of late!
I was even having the sort of same feeling about AI voices as of late, mostly with the release of CiFlower, but now that producers are showing what they have to offer maybe it won’t be so bad after all, plus I know that I’m at least not alone with what I thought, so keep up the good work 🎉🎉🎉
Hello hi, im grateful for this video since its edited well and its catchy But i want to put in some details about how CeVio works.
It uses HMM like Vocaloid 1 but onwards vocaloid stopped using it. HMM observes some recordings and recreates them while vocaloid 2 onwards takes the recordings and stretches them like UTAU. Cevio with ONE added auto-tuning to non AI vbs and about the tuning this isn't like we expect, it just averages the pitch between silence and vowels. It's mostly extracted from modulation so you have to tune AI voices the same way like you do with HMM but the difference is that AI will re-render the note more naturallay this means it extracts data like strength-breath-accent based on the song data it has a similar pitch bend to what you did in theory.
edit: I wanna add about backround noise. HMM produces more backround noise since it recreates a whole sound through software V1 and CevioCS have a similar air like engine noise.
V2 onwards and UTAU produce backround noise is the sample needs to be stretched - vocaloid uses exact cuts and loops so noise woudnt be created while UTAU will produce a metallic sound usually the example you showed didn't have engine noise, the backround noise is actually the formants that are peaking and resonating in the recordings so that's a microphone problem, they are usually heard louder in utau than the recording cuz of compression.
Woah that's interesting. I apologise for the lack of a clear explanation and my mistakes in the video, I don't know much about this sort of stuff so this is quite the insight. Thank you for your comment!
fuckin' love KAFU's voice. one day, one day.
With the announcement of POPY and ROSE, this is a good way to learn more about CeVIO, thanks a lot!
hello, fellow BanG Dreamers 👋
Wow, this is amazing! The production quality is of course through the roof, and I learned lots of things I didn’t know about CeVIO before! Do you think you could do one on Synth V too? I think they get some really impressive results, especially with Solaria’s recent release. You got a new subscriber!
Also one thing I’ve heard people complain about with CeVIO is that they can’t sing as fast as other synths, even other AI synths like Synth V AI. Past a certain speed their engine noise and failure to properly enunciate syllables makes them impossible to understand. Also this is my opinion, but I think they have a tendency to sing flat as well. This can be fixed of course, and for some voice banks it works in their favor (most notably KAFU) but these issues contributed, I think, to issues in other voicebanks, including IA AI’s lukewarm reception
Thank you for your kind words and new information! Regarding Synth V, I might make a video about it someday but I don't have plans for it anytime soon
8:45 this very much. i love how kafu sounds, but yeah it does start to sound very similar across all artists… basically kafu becoming more like a digital utaite rather than a voicebank. which isn’t a bad thing! but it’s different!
We were really living in the best era of technology, haha! Anyway, I've got the insight of it, a really helpful and great video, thanks mate!!
I like the fact that CeVIO AI and SynthV have a higher quality and realistic voices. However, I none of the songs makes with this voices got attached in my brain like vocaloid songs.
Maybe it's because VOCALOID voices are more catchy and more addictive to listen to.
The editing is impressive.
So I was wrong for 4 years... I thought that CeVIO is a part of vocaloids... Like the way of tuning or smth makes them special, not-like-the-other-vocaloids
This deserves a million views bro
damn this video was amazing as someone that has a big interest in how vocaloid software is made i have learned alot about this good job man
👍
Cevio sounds very realisticly when used right. But even producers like ATOLS can't get rid of the "nose tone". The singers sound like they have a stuffed nose and sing through it.
This is why I am very puzzled if I rather have a clear robotic voice. Or a stuffed nose realistic voice.
Touhou 🤧💖
Thanks for the info! I've left the vsynth community years ago but i like to keep up with updates now and then.
Also! CeVIO is largely based on Sinsy, (and CeVIO *is* basically commercial Sinsy), so if you wanna see the development of CeVIO, you should look at Sinsy's first.
Wait really???-- that makes so much sense.
underrated channel
do hope you fetch more success
Great video! It's great to know how this works
and now we have cevio pro!
When the Dave fucks up the script and had to redo the entirety of it in a single day
@Hatsune Kiko yes
wow davee! great video! i have "liked" and "shared"!
Wow, great video
This so good!
Great video
great video 🤠
But what does "CeVIO" stand for, is it an abbreviation?
I knew it the music sounded like touhou
5:11 thats actually Sasara
LETS GO
Chevio ai feels like kyaami tuning
Wait, it's not pronounced see - vee - oh??
Por fin lo encontré :'v
Wait wait wait
It’s not pronounced See-vee-oh
Yes mom I am studying. I am studying ... CeVIO 😳
Please do a video on synthV
SynthV is pretty interesting but I won't be making a video on it anytime soon. I might make it one day though!
@@daveejoneseywill you make u video on the different Vocaloids
@@vaityv8967 probably, I don't know what the next video will be though
@@daveejonesey ok ,have a good night or day I don’t know ur time zone
is UTAU the only voice synthesizer that allows users to create voice banks so far?
As far as I know, yes, only UTAU has that feature
am i the only one who frequently confuses synthv and cevio? the release so many same voicebanks xD anyway this video is awesome
Me lleva el demonio quiero cevio pero pobreza epic
touhou music!??!?!!?
Hewwoo OwO
La grasa logo :VVVVVV
With all of these other synthesizers popping up, it would be safe to assume (and unfortunately) admit that Vocaloid is essentially dead.
I was so excited when you said machine learning earlier in the video but then you switched to calling things AI which doesn't exist 😭 I understand calling it Cevio AI as that's the name of the program even if the last part of the name is untrue.
AI would be things like HAL 9000 from 2001 a Space Odyssey, Cortana from Halo, or Glados along with the cores and turrets from Portal. The only reason companies call what we have now AI is because it rolls off the tongue better and is catchier but they're still calling a plant a human.
SDLG
CeVIO just sounds wayyy too nasally for me .
😲
In my opinion I really dislike CeVIO…I do think it is a step down from before.
Most voices like kafu and also hearing Flowers demo I really dislike them.
And now with all these Producers switching over to CeVIO I’m starting to get a little borde and starting to really hate CeVIO more
ojalá saber inglés
I love kafu
bro
:V
Funny shitcomment :u I'm dyslexic and Spanish and always pronounce this software "Ceivo" 😅, I'm an idiot