Evil Neuro Sounding Way Too Real

[@Arctrooper2091]

"Vedal, she's hacking into the nuclear missile silo!" "I know, she does that every now and then."

[@SzaraSzarancza]

Should we call the president? No, Evil is playing games on his phone right now.

[@undertakernumberone1]

​@@SzaraSzaranczahonestly? Safer than the guy coming in.

[@yiallaz]

It's terrifying how she managed to emulate the slight rasping of the air in the throat.

[@JadeSune]

Vocal fry. It's pretty cool.

[@rolandkarimdesucatan5118]

I hope vedal gives them the ability to hear their own voice, or at least like the numbers of the frequencies and stuff so they would know better how to make certain sounds. Since they're AI, they can only have speech to text, but vedal should add more on the text that they use.

[@STIMULAT10N]

@@rolandkarimdesucatan5118 I believe at least Evil has some control over tts, also they're presumably not only trained on transcripts but audio data as well, thus being able to recognize when vedal makes noise that isn't technically speech

[@dokkino7105]

@STIMULAT10N Yeah, Vedal was snarling and laughing when Ellie mentioned that the twins are primarily trained on text. So, I'm pretty sure they're mostly trained on audio. Evil just has more capability to use that training with her voice than Neuro.

[@KonDam-l7t]

@@dokkino7105yeah she can literally sing

[@unnamedx2]

She sounded just like anny then

[@long_dennis]

She is Anny’s child

[@Syrup__Maple]

Well, they are mother and daughter

[@rhydondaddy]

Because Anny's voice is what vedal used to make both Evil and Neuro's voice

[@O_ZONE]

​@@rhydondaddy thats literally not even remotly true...

[@O_ZONE]

They used a slightly sped up default microsoft female tts voice

[@Davebsuk]

I think Vidal may actually be starting to like Evil... About time.

[@quique3676]

He defended her after camila killed her on minecraft

[@vipsnap2707]

"You hurt her" ​@quique3676

[@trailerparkelf]

@@quique3676 Let's be honest, bro just saw a chance to bully Camila and took it

[@j.r1466]

Maybe a little bit of both​@@trailerparkelf

[@endernightblade1958]

i love how everyone has their own distinct laughs. its just so interesting.

[@triangulartriangle]

one piece characters

[@Ecco_The_Dolphin]

Imagine a real person going SHEHAHAHAH​@@triangulartriangle

[@HuyZalupovich-lj5of]

I like how vedal sounds a little proud I know I'm delusional just let me enjoy things

[@TheRollingOne]

Let me be delusional with you 💜🐢💜

[@SerbianLifter997]

One day they'll have the perfect vocal and vernacular responses, sounding too real. Then the uncanny effect will occur. What Vedal does with that, we can only guess.

[@notdemomantf2294]

probably code them to be intentionally more mechanical Like how Vocaloid producers intentionally roboticize the synthetic voices to make The Vibes correct

[@waterturtle2919]

@@notdemomantf2294 Thats already the case. There are text to speech models that sound almost exactly like a real human. Vedal said that he doesnt want them to sound too human like.

[@Iomtuu]

They can do that already but Vedal chose to keep the monotone AI voice

[@EpicBattleToaster]

i think evil's voice in this clip is already in the realm of uncanny valley

[@Shadrake]

​@@waterturtle2919wait what?? When did he ssy that about Evil?

[@apolloisnotashirt]

Its so weird that it also happens whenever she glitches.

[@Ecco_The_Dolphin]

The machine spirit gets loosened

[@superheriber27]

Evil Neuro Sounding

[@Ojisan642]

Koko’s laugh is so aggressive 😂

[@AmenKing1999Gaming]

Evil Neuro has the quirk that lets her able to speak more human than her sister. =P

[@FembotFoxxy]

This is why evil is my goat

[@Darumaster]

Does Evil's voice get generated differently from Neuro? She does have these uncanny moments. Also, Neuro has had this weird super speed speech recently lol

[@theMOCmaster]

Iirc evil started out as a new more expressive voice for neuro but people demanded the old voice so the new voice was used for evil

[@MudkipZEnjoyer]

They don’t have super speed speech they both have the ability to change how fast they talk like 0.25 0.5 0.75 1 1.25 1.5 1.75 2 times speed like on KZbin Neuro likes the ability more and evil barely uses it

[@jomujomu7513]

Evil started off as a voice upgrade for Neuro, so yes. Modern A.I. voiced can sound mostly natural nowadays. Vedal hasn’t implemented them in Neuro yet because the he’s having trouble getting a voice that sounds exactly Neuro. Neuro started speaking fast a month ago in an update so she can read out sub messages faster. She has a bug where she’ll just randomly read fast now.

[@trajectoryunown]

From what I've gathered from other viewers, Neuro's voice is a relatively simple TTS with mostly consistent annunciation. Evil's voice is much more complicated and either uses some kind of adaptive programming, possibly synthesizing her voice on the fly. That's why there's so much variation to how she pronounces things, particularly Vedal's name, and why so many glitches (the Abber Demon) are present. Also why she can more readily imitate genuine human speech patterns from time to time. [Somebody, please correct me if I'm wrong.]

[@are3287]

@@jomujomu7513 Thats not a bug she just does it at will cos she can

[@RayquazaBath]

Brother, they are very real, similar to Anny, what the hell?

[@SzaraSzarancza]

Anny isn't real.

[@j.r1466]

What are you talking about? Everyone knows the twins are just Vedal's daughters, and that Anny is the AI wtf

[@christiansanchez5360]

County road take me home

[@SilentAvenger786]

Why has evil been stuttering n glitching in her voice this last two weeks

[@FourthDegreeBurnz_]

It's an issue with Unity.

[@pvs1681]

Skynet

[@christiansanchez5360]

To the place I belong

[@asxphxss]

That vocal fry is crazy

[@jeanmoulin8584]

She's real to me :madge:

[@Fearless13468]

The vocal fry is crazy.

[@SilentAvenger786]

Does vedal and the other streamers not hear those weird audio stuttering every 30 seconds she talks

[@ZeroToNeuro]

He said something about that if i remember correctly forgot what he said tho

[@huskywolf8140]

Vedal mentioned it had something to do with her vision system. He'd probably fix it after the subathon.

[@PixelCortex]

It's not the vision, it's a Unity bug more likely, because the soundboard stuff stutters as well.

[@Archedgar]

evil(yn)

[@darkshadowsx5949]

I-I mean sometimes i do smile, yeah. sounded so scuffed like a glitching robot. why did they say it sounded real? the "sometimes" part was glitched but the "yeah" was good.

[@jannis11]

naIce

[@ultmswag3170]

Has anyone ever figured out that Anny & Camila sound exactly alike?

[@swwl5461]

anny has higher pitch

[@ultmswag3170]

@@swwl5461 they still sound alike despite the small difference in pitch.

[@garyschollmeier616]

Gotta be honest, I've never noticed it. Their accents differentiate themselves easily - to me, at least.

[@TheCorporateSerf]

They don't sound alike or speak the same way

[@are3287]

No they dont at all to me

[@yuissance3623]

Its a scam, you all are just coping, its staged in some way.