SynthV Eleanor Forte AI announced by Volor

[Scarlet Illusion]

You know, as tack and busy as it looks, I actually really loved it. The text in the background makes me think what a vocal synth box designed by Gucci would look like, lol. Loved every bit of tackiness right up until I saw her hands were tied in the ribbon and then had flashbacks to creepy bondage-type artwork I've stumbled upon by accident when I was naive and thought I could look up a character on DA without seeing creepy stuff, haha.

Other than that not-so-lovely detail, it's very nice! I notice Ellie's face looks very expressive, but I don't think the artist changed? To me it looks like whoever the artist is improved at facial expressions!

 

[mobius017]

Offhand, I really like the box's clean, bright color scheme. The repeated name across the box is a little busy, but I think it might've been a more common design element 1-3 decades ago, so it strikes me as kind of a retro choice (Which fits with Ellie's persona, actually, IIRC--she likes writing/pens/maybe somewhat older things?). After awhile, it sort of fades into the background. I wouldn't be surprised if it were a placeholder for something else later.

Only the ribbons around her arms/legs perturb me. I'm not sure if I jump to the bondage association or not--I feel like I've seen other product containers where characters pointlessly had their arms tied like that, maybe just for the sake of having a reason to put a ribbon/bow in that exact spot for artistic reasons (e.g., the color balance/addition of that shape/something to fill empty space)? But I was thinking that before squinting at the bottom and seeing that her legs are maybe tied together, too. It seems pointless to put bondage elements on a vocal synth box--the rest of the box doesn't scream "sex," it has nothing to do with the product, and I wouldn't guess it makes the product more appealing to the majority of potential purchasers. Putting that association on the package might be flirting dangerously with Chinese content laws, too (Ellie/Synth V do hail from China, don't they?). It wouldn't be the first time people have lived dangerously with stuff like that, though.

Other than the ribbons, offhand only the ampersand on the spine jumps out at me. I feel the pain of the poor person who got the character's name wrong....

 

[ikaroll]

Ellie/Synth V do hail from China

I think Dreamtonics is a taiwanese company, someone feel free to correct me on that

her art/box is pretty, i won’t deny that, but personally, i find the ribbon tie stuff out-of-place for her design (and like others have added, a bit creepy)

 

[xuu]

Animen are a company based in Taiwan with a Japanese branch, Dreamtonics are a Japanese company. I feel like I'm saying it a lot but this box is a placeholder and will be different to what her final box looks like, it's just something to show "yes, she will have a physical version".

 

[kozet]

Got an e-mail from Team VOLOR.

Dear customer,

Thank you for joining Eleanor Forte AI's first-round early-bird pre-order. After correspondence with Dreamtonics on Eleanor Forte's development progress, we have some new information.

Regarding Eleanor Forte AI's progress, we expect that the database will be completed by the end of October, meaning VOLOR can distribute the database to first-round customers in November 2021. We were pleased to hear from Dreamtonics that the recording quality of the database is excellent.

The second announcement is that VOLOR will run a limited Eleanor Forte AI package version pre-order campaign starting October 9th. Customers who have already joined in the first round will be able to add the package version to their order for a discounted price. Please look forward to an email when the campaign starts and check our Twitter account @OfficialVolor for any updates.

VOLOR Team

Guess we'll have to wait some more but at least we have some more info.

 

[Cluemily]

On Eleanor's product page for her physical box the product image has changed to a different box design it seems. This one looks more like the more standard box design a lot of the voicebank physicals have, with the dark spine (and the thicker case, possibly?).
(Page still states "Physical box design picture only for reference and can be different.", of course.)

 

[aru ii]

Her demo is supposed to be out already, but I can’t find it anywhere. Does anyone know where to find it?

 

[___]

Her demo is supposed to be out already, but I can’t find it anywhere. Does anyone know where to find it?

The annoucement was hard for me to read with my learning disorder too, but if I'm not mistaken, it's coming out tomorrow, it's not out yet.

 

[Rylitah]

They've changed the premiere time - it's out now. This is the voice focus version, which Crusher said in the chat is the song without percussion. Normal version of the song is available for download in the description.

 

[WyndReed]

Even with the normal version she sounds really bored to me. I was hoping for more realism.

 

[peaches2217]

I love her tone, but... yeah, it leaves a LOT to be desired, I think.

 

[morrysillusion]

yeah honestly if i wasnt told it was an AI bank i probably wouldnt have thought it for a moment, as it didnt strike me in the same way as some of the already existing AI banks in terms of default realism. not bad exactly, i love her tone, and considering i only own her old standard lite version i dont regret my on sale pre order as i still feel its an upgrade from that personally, but it feels a bit lacking compared to some of the other AI voices.... im sure people will tune her in lovely ways but in this default, less tuned state shes just okay to me.

 

[inactive]

The voice sounds clean and not scratchy, so I'm satisfied. Besides, I don't do the wibbly-wobbly pitch bend thingies that are so common in modern pop music, which means that some of the artifacts heard in the above sample probably won't be an issue for me.

Although they could've put some effort into de-essing...

On the other hand, sibilance is a normal part of speech/singing, so perhaps they wanted to demonstrate that Eleanor Forte AI's sibilant sounds can be just as annoying as a non-synth vocalist's sibilant sounds.

 

[___]

They got that part right!

 

[lIlI]

She definitely sounds very tired/emotionally flat in that clip, but I do like her tone and singing style the most out of all the Synth V AI voicebanks so far. I wonder how much tuning effort would be needed to give her more energy?

Concept I'd be very interested to see: An AI based on a very distinct emotion. Say they only used angry songs for training, could an angry singing style be replicated? This would be a cool and useful way of doing appends. In fact, I think a voicebank linked to a specific emotion would sell really well! Sure, it would be limited to songs with that mood, but it would give a powerful result without requiring much tuning: whereas neutral voicebanks always need editing.

 

[mahalisyarifuddin]

Concept I'd be very interested to see: An AI based on a very distinct emotion. Say they only used angry songs for training, could an angry singing style be replicated? This would be a cool and useful way of doing appends. In fact, I think a voicebank linked to a specific emotion would sell really well! Sure, it would be limited to songs with that mood, but it would give a powerful result without requiring much tuning: whereas neutral voicebanks always need editing.

Said concept would theoretically works.

But, if you think about it, wouldn't it makes the Tension parameter—one of the main selling point of the engine—essentially useless and redundant? Though we are still waiting for the Glottal effects redesign

Sure it would be easier if we want specific emotion, but it will sacrifice versatility.

We already experienced it, especially in Vocaloid (we don't talk about Utau because people can do whatever the hell they want). Appends have jarring qualities each other, and I yet to see Appends that can works seamlessly, and that's why V4 introduced XSY—which has been killed in V5.

Not to mention it would be expensive to produce.

And yeah, less is more, especially if you can make the most with less. And you know what? That's Dreamtonics' design philosophy.

 

[peaches2217]

Appends have jarring qualities each other, and I yet to see Appends that can works seamlessly, and that's why V4 introduced XSY—which has been killed in V5.

To be fair, V5 introduced emotion tools that provide similar results with greater predictability and reliability than XSY, which was notoriously buggy and offered greatly fluctuating quality if results (and I say this as someone who used the crap outta XSY when I mained V4). But that said, Appends work well for Vocaloid and its setup; a large part of the appeal of SynthV is that, even lacking Appends, you can get drastically different results with the same VBs thanks to the parameters.

On the other hand, I feel like that quality has been mitigated to some extent. The parameters don’t offer the same drastic results as they did in Gen1, and SynthV banks are becoming near-exclusively AI, which don’t offer much flexibility at all by nature. With that in mind I don’t think making an AI bank with a targeted emotion would be a bad idea at all. Given how bland and samey a lot of them are starting to sound, I actually think it would be a great idea; they wouldn’t be any more limited than SynthV AI banks already are, but they would give us more options.

 

[blackout]

This demo sounds um... Real barebones I'm just gonna say. Think I'll wait for a bit of a better usage to rly see how I feel about her

 

[Blue Of Mind]

I dunno why, but Ellie AI reminds me of how V1 and early V2 Engloids would sound without the right tuning - a bit flat and lacking emotion. I agree with everyone who's said they like her tone, but I'm really wondering now if you can get more power out of the AI bank (and I'm aware Ellie Standard already sounds soft to begin with).

 

[lIlI]

A new sample!


Ooh, her voice has a little more life here, and the differences to her Std are more obvious. I definitely prefer her AI over her original voicebank: it's partly due to voicetype preference. She sounds like a normal adult here, whereas her previous voicebank had a stronger moe performance, with more engine noise. I would understand if people who prefer a cuter tone missed her original.

The main downside for me is that, thanks to her softness, she sounds quite similar to Anri. Eleanor has a more formal, theatrical tone, but they aren't as distinct from each other as Eleanor was to Maki. If Anri and Eleanor's voices were suited to different genres I would have been more likely to consider buying both. Though good for my wallet, it's a shame to see a spot taken that could have gone to a voicetype we don't have yet.