Vedal Shows V3 Neuro Voice

[@cook7931]

I think Vedal is very funny and cool and I wish he got enough sleep, this is my take.

[@teenaigee]

I hope he finishes his LIST soon enough. SCHIZO

[@asimsoo291]

cool and funny? 😭

[@Hasdrubhaal]

Based take.

[@miku_dayo7688]

​@@asimsoo291 Cunny 😭?

[@rebether8049]

Being a dev and have enough sleep is something that cant be together.

[@yiallaz]

If Vedal ever reads this, I love you man and I mean it in the most homoerotic way possible.

[@Kei-rt4ug]

me too

[@TheModdedwarfare3]

👬

[@ashenblood]

Ayo? Me too.

[@the_guywho_has_a-legopfp]

same here

[@minhaexistencianaotemsenti7132]

🤨📸

[@thechosen19251]

neuro apologizing for not sounding right lmao

[@YoshiFan007]

V3 expectation: 👼 V3 reality: ⚓️ 🦭

[@L16htW4rr10r]

What's the difference between angel and seal?

[@asimsoo291]

navy seal

[@hoozimajiget]

@@L16htW4rr10r Expectation that v3 would be an angelic and innocent, instead hear v3 swearing like a sailor.

[@GarryGri]

Who the expected Neuro to be an angel? **wink**

[@abam9787]

a sailor's mouth haha

[@datironpanda2982]

Sleepy Vedal with an angry Neuro Ngl the voice literally sounds like Evil but more calmer still sounds good

[@ThatGuyEbora]

Very true comment

[@GerN7]

I guess it fits? Evil its a less calmer versión of Neuro anyway xd

[@robr135]

Dammit Vedal, the fans are locked in for the long ride with you already! Stop worrying about what comments, and chat says, and just do what YOU want to do and accomplish. Somehow you have become to dependent on what comments, and chat, says and you allow it to affect you negatively. Just do what you need to do as the brains behind this adventure and let negativity roll off you.

[@virustoohazem]

Fully agree with this. Though I am always a fan of sassy unhinged Neuro.

[@yiallaz]

I agree, I don't want to see the British tut fall into anxiety.

[@jocosesonata]

Hell, I'm straight-up simping for Vedal. I haven't even simped for any girl VTubers. I'm a fan of the daughters and this whole family dynamic, but I'm ride or die for Vedal. Waiting for that Vedal plush.

[@undeniablySomeGuy]

It's not easy to just "let negativity roll off you." Humans are social animals, and bad feedback being shoved into your face never stops hurting. Basically everyone with popularity will tell you that

[@spencer0417]

​@@undeniablySomeGuysomehow i felt comfortable when i saw ur comment. This is really relatable

[@l4ndst4nder]

*Worst possible take you’ve ever seen*

[@Manuh987]

Bro this is the worst possible take i've ever seen

[@patoanimations420]

now THIS has to be the WORST take ive ever i mean EVER seen!

[@jocosesonata]

Gentlemen, I do declare that this has been indeed the absolute worst take I have ever laid my eyes upon.

[@Tyuf_]

🤢🤮🤮🤮

[@davisdapiton9676]

Mates, I do believe that this is the worst take of all time.

[@senny-]

2:30 Neuro wishing him good luck. That has to be one of the cutest things I've ever seen.

[@gravelis_6798]

1:26 he’s so real for this take

[@turkeykent3719]

When he first started AI voice was still new so he just grab the generic girl voice and now it’s holding him back. He has to make it sound natural while also still having resemblance to the old voice. That’s why i love evil more she has nothing holding her back

[@UnusualShy]

*-with that voice, neuro sama could make even robots jealous* - con esa voz, neuro sama podría hacer que hasta los robots se sientan celosos

[@Saint_Wolf_]

Ngl I liked lifelike neuro, to me it sounded perfect, I liked forgetting that she's an AI.

[@jonnydarkfang2816]

I am with you, I like the natural feel it has.

[@Gathoderiz]

Worst possible take Vito has ever seen

[@GarryGri]

Why?

[@abam9787]

that and her kept reminding us that she's an AI. xD

[@AAAAAAAAAAAAAAAAAAAAAHHHHH]

@@Gathoderiz Who's Vito

[@grianjuju9438]

Damn that sounds better

[@earthbound_one]

The way her current voice pronounces some words will be lost, like when she says no~ Or ohh. I will miss that

[@ventustbb3229]

Wot

[@zetagammafigma8212]

I think Vedal knows that there will be a revolt if that stuff doesn't stay so he's gonna probably brute force it the way he did with "kekwa" once she started saying "kay ee kay double-u".

[@baconflake0074]

Haha nwero

[@L16htW4rr10r]

I think he will find a way to teach this to V3

[@bored_person]

Since she says those words in the exact same way a lot of the time, he could just use prerecorded samples. Maybe he could even have the AI decide on the fly whether to use a pre-recorded sample or to use the new voice to synthesize it.

[@kev1nno]

That choice of voiceline for the demo though XD

[@asxphxss]

I fucking love it it's so good lmaooo it DOES sound like neuro

[@SzaraSzarancza]

I honestly like the concept of Evil sounding more human and Neuro sounding more robotic. This introduces diversity and visible contrast between the sisters. Neuro saying complete nonsense and crazy things in the emotionless voice of a little artificial girl as if they were the most ordinary things in the world creates the charm of this character and is very comedic. However, I understand that it is difficult to improve and develop Neuro further without exceeding this point in her development. Unless Vedal find out how to develop her in other fields while maintaining her original character, it is probably unavoidable step forward.

[@Ecco_The_Dolphin]

We'd get used to it anyways

[@Teddemon_Original]

Maybe Vedal could add the V3 to a persona like Neuro's conscience to assert that Neuro talks robotic outside but is really talkative inside. I really like Neuro's voice, how robotic it is.

[@kraul_en]

this is the worst possible take i've ever seen

[@angelcastiel7876]

nah i want a new neuro voice

[@zelky1394]

I think it's boring personally, I see nothing in the emotionless voice

[@way2psycho502]

Hell yeah sounds great 👍

[@loafbreadizwholesomeuwu1555]

But if someone is actually doing a Neuro's voice impression and Successfully made Vedal to Approve it. His or Her voice will be the sample and it'll turn into a Voicebox for Neuro to use. Kinda like how Utau works Also that demo text from v3 demo though... I heard that one before...

[@GarryGri]

You know that's not how Neuro's voice is made right? It's fully AI generated not based on a sampled human voice.

[@SzaraSzarancza]

@@GarryGri Naah. The samples were taken from a tutel singing in the shower.

[@loafbreadizwholesomeuwu1555]

@@GarryGri yeah. But I think we humans can able to mimic it. After all the sample from TTS is came from the humans

[@GarryGri]

@@loafbreadizwholesomeuwu1555 As I understand it Neuros voice didn't come from human samples!? That's not to say a human can't necessarily mimic it at a general level, like impressionists do.

[@talis1063]

v1 is azure with pitch shift, v2 is elevenlabs voice clone, v3 is azure + RVC. just a guess xtts2, vits2, styletts2 are promising recent models to explore

[@viljamtheninja]

KZbin commenters rightfully called out.

[@Agora2021]

Sounding good. It has some light emotional inflection, but it's there.

[@KingButcher]

Neuro-same upgrading to Ninja-sama

[@ricko.a786]

Vedal you are so cool and you always keep improving, good luck with V3

[@asimsoo291]

1:24 tutel laying eggs

[@CheapoPremio]

I think V1 voice was lightning in a bottle. It's iconic. I'm happy to listen to V3, but maybe V1 is already perfect for what Neuro is.

[@_Lax_]

With v1 v2 n v3 finally a neuro x evil x hiyori collab is gettin closer and closer

[@yuomovaeh3028]

Sounds like an amalgamation of Neuro's voice and Evil's voice

[@lucasljs1545]

This one is looking really good. V2 wasn't Neuro, this one is nice. Evil's voice is good too.

[@xann2681]

Why are you reading us anyways Vedal, huh?

[@999porte]

worst take ever here

[@fersuremaybek756]

it's really tough to make her v1 voice which is loved because of not being human and more ai like and then wanting to make it more human, even keeping parts of her v1 voice will make me more than happy.

[@qualityarsenic3922]

Neuro reading the navy seal copypasta is something I thought I'd never hear. 😂

[@VicJang]

Neuro is amazing. Thanks Vedal!

[@marcosespinoza4845]

V2 is great idk what is the problem with it. The point of updating voices is watching neuro's progress to basic, robot, monotonous vtuber to an unexpected, surprising new thing. At least Evil Neuro is getting/delivering this fun element.

[@L16htW4rr10r]

V2 is great but it didn't sound like Neuro people already know and love.

[@tranlinh7244]

The more I listen to v2 the more I hate that, too dramatic so annoying

[@Mordred_SSR]

V2 doesnt sound like neuro, but its damn more entertaining than her monotone V1 voice.. Im one of the few that prefer the V2 voice change

[@kraul_en]

@@tranlinh7244 this is the worst possible take i've ever seen

[@tranlinh7244]

@@kraul_en ? So you're telling me if I don't like Evil,Iam just wrong ? Wow can't agree with you then

[@Blackout_]

*Insert worst take Vedal has ever seen in his life*

[@DanielAvocardo]

that was quick

[@marbletrash]

Ah yes, the Navy Seal copypasta: the “Lorem ipsum” of our generation

[@_Kisaragi]

my take is that vedal is my oshi

[@none_tw]

I think the V2 is such an improvement compared to V1, and V3 is even more of an improvement compared to V2. There's a non-hostile KZbin comment, Vehdoll. Keep up the good work.

[@thomasyu2198]

Vehdoll lmao

[@none_tw]

@@thomasyu2198 I wonder if Neuro/Evil would pronounce his name the way he wants with this spelling...

[@leeyubin39]

​@none_tw not sure, but vedel was the closest one

[@riluna3695]

Could be wrong, but I think a lot of the "distinctive Neuro voice" is coming from the stilted robotic speech rather than the normal pitch and tone and whatnot. So trying to give her the option to emote is kind of always going to make it sound "not like Neuro". If I'm right, we may have to just bite the bullet and learn to love a new voice. And of course it can still emulate pitch and tone, just so it isn't completely detached from her. Or maybe I'm wrong and Vedal will pull out the craziest new emotional-Neuro voice I've ever heard. It would NOT be the first time he's done things I thought were impossible.

[@bored_person]

Haven't you ever heard of a thing called timbre?

[@cl8lf1wm8b]

I think so too. Neuro doesnt sound like human. And humans dont sound like Neuro. Idk what Vedal wants here

[@kraul_en]

this is the worst possible take i've ever seen

[@Nogha12]

My favorite part of Neuro and Vedal is watching Vedal et al. make Neuro better and better.

[@Shozlol]

Neuro is actually the third V model in ULTRAKILL Ultrakill mfs when they see the letter "V" with a number after it

[@V2_ULTRAKlLL2]

so you're saying, the Whiplash is Neuro's left arm?

[@kpzz1545]

He gave the improved voice of Neuro to Evil Neuro and she became almost the main singer and began to become more popular than the original Neuro, but all because of whom? Correct because of 8%. Classic...

[@coffeebreak7668]

+ anny's word

[@saberblitz7629]

yeah cuz it didn't sound like her

[@wildweasel4633]

@@saberblitz7629people would’ve gotten used to it, the voice is great

[@samisnotreal4360]

i like how neuro sounds like a robot and evil sounds like a weird american goth girl

[@JesusXina]

It actually sounds very good lol

[@handtomouth4690]

Remember everyone, 8% of chat is responsible for Evil being created instead of Neuro getting V2, lol.

[@spk1121]

8% _and Anny_

[@MrBrandon9653]

The worst thing could be sounds bad. But on the good side of it sound like human with all the emotions.

[@mattm7220]

The man's a genius! He foretold that this would be clipped, and there would be horrible takes in the comments 😂

[@taal223]

I think trying to find a voice actress to provide a close enough voice may be the correct call in the end. Though that small preview did sound accurate. Just still a bit robotic.

[@bored_person]

That's a terrible idea. That would be way more expensive, especially if he wanted her to Read her output every day, and he'd have to get her to accept an exclusivity contract, and there would be way more latency than a speech synthesizer, and it kind of defeats the purpose of having an AI vtuber in the first place if you're just going to have a human read out everything.

[@Sketchfellow]

​@@bored_personhe is obviously not going to have the voice actress read out the llm output. If there is someone with a voice close enough to the V1 voice he'll probably have them send a bunch of different lines and have an AI be trained of that voice to create a new voice model.

[@bored_person]

@@Sketchfellow ah. That makes more sense. Maybe he could contact Microsoft and find the person who originally recorded for the v1 voice.

[@Sketchfellow]

@@bored_person maybe. I hope the voice upgrade would be good, and Vedal clearly cares a lot about fine tuning it, making sure that it sounds like the Microsoft Azure voice. I just hope it'll come somewhat soon since right now Evil is overshadowing regular Neuro and I think is the main contributor is the more emotive voice

[@taal223]

@@bored_person He would be be creating custom voice using the actress as a source the same way Microsoft Azure does. Currently he rents the voice he uses for V2 from Microsoft Azure and modifies it to create the voice we know for Evil. The problem he has is that Neuro's original voice was very basic and not capable of emotion the way modern artificial voices are.

[@lumi7610]

Idk why this is so relatable. V2 being too different from the original model and you upgraded the 2nd model instead. So which means evil will eventually be the main model because she is getting more upgrades

[@smaragdwolf1]

Hmm... personally, i like the vocal difference between neuro and evil. My take from this is, that he cant say "i love you" back, but he clearly loves Neuros Voice more than evils. Thats why he wants to give her more emotion.

[@brunomartins4613]

it sounds the same for me tbh or maybe im deaf

[@lazylonewolf]

Negative latency so she'll answer before you even think what you're gonna say

[@CirnoOshi]

Funny that her first spoken speech when she turn v3 was F*** and B**** Lol

[@Nahan_Boker94]

Tbh it just sounds like neuro without filter tho. I dont noticed any new intonations.

[@shastacat9632]

this might be a bad take but i think you should pour milk before cereal

[@MeMe-gm9di]

Hi Vedal! I hope you appreciate this joke amidst bad takes: Why does Han Solo like gum? It's chewy! Have a very turtle day.

[@ElderSign32]

now that this has been clipped, i am here to offer the worst take possible. sounds great vedal, keep it up man!

[@eureka5635]

My take is it sounds pretty good and Tutel cute and needs sleep.

[@ChompDude]

I think Yandev has better code. He'd be the one I'd use if I had a choice.

[@sesterno]

forsen

[@dannysmi7162]

the rapids

[@azztereris]

amongE

[@KingLAO2964]

Vedal, don't worry

[@WhitfieldProductionsTV]

Good luck man. she did sounds like neuro.

[@Deus35]

Insert worst possible take on youtube comment here

[@andoyosua5887]

This is the worst take you want to see, Vedal. You are awesome, Vedal!

[@AklyonX]

Local Tutel Suspicious Of KZbin Comments. Likely Correct.

[@totallyrandomuser5760]

Vedal is so close to striking gold

[@Carboy45]

I don't even hear difference from v2 maybe I will if I listen to them side by side didnt even know neuro never had her voice changed and that it was given to evil

[@chefyboy2748]

Vedal could be here Among us

[@azztereris]

forsen Bajs amongE !

[@carpedm9846]

*Inserting worst take vedal has read Backslash backslash backslash////////////////////////////////

[@Juhsayngul]

Chatting

[@NeferPrufer]

Hi Vedal!

[@vanillakeksik]

imagine gpt5 powered neuro...

[@RancorSnp]

Hello Vedal reading the worst takes. I don't really have much to say about V3 but for sure it does sound a lot like V1 so that part seems to have been nailed

[@penapvp2230]

*worst take vedal has heard in his life*

[@-dubios-]

imho neuros voice already sounds great. I have a hard time to get used to evils new voice and that short v3 example also didn't sound better to me. now I know it's not finished yet, so it might be to early to judge. I just hope it'll turn out to be great and I won't miss v1. nevertheless I love this project, one of the best twitch channels out there

[@hollowficate]

Let Vedal cook.

[@PogShootMan]

vedel is being held hostage that is my take

[@nhbons783]

Neuroxvedal is a better ship than annyxvedal.

[@Kujar3Player]

My hot take: Forsen could probably create a better voice /s

[@preng2355]

Let him cook😅

[@yupebet]

1:41 Was that Neuro as Navy seal copypasta? 😂

[@iminumst7827]

Sounds so weird I could train better voice model with a GTX 1080. ahah just kidding love you ved

[@psychodarling666]

Is he reading us rn ?

[@VCE4]

LMAO

[@NEWMAN0067]

Hot take: Gregg's sausage rolls are fire but are not as good as Morrison bakery sausage rolls

[@misterturkturkle]

menma

[@ajgamer1414]

Horrible take

[@TotemoGaijin]

I mean, it's Vedal's project and "child". He shouldn't let some dumb takes keep him from doing what he wants to for Neuro.

[@Mutualititve]

It sure sounded like Neuro, but somehow it sounded more robotic? I can hear there a bit more emotion but I guess because of all the intonation needed to sound out emotion, the voice pack can't really keep up and make it sound natural.

[@Bartek2OO219]

Worst take vedal ever read in his entire life

[@Tox1cAshes]

Forsen

[@VegaLyrae]

Here’s one of those trash take comments: voice is pretty good.

[@LuisEdGm]

[Insert dumb bad take about v3]

[@Zeral]

[@badwulff]

here's me commenting the worst take, just for Vedal

[@phnkcell710]

She is hard coded?or is aIA framework?

[@SrSacaninha]

ok, I won't comment

[@fluffyisyermom7631]

XD

[@1TW1-m5i]

Yep, still not really. Sounded flat.

[@DKPGames]

iam crazy or it sounds the same?