SynthV Eleanor Forte AI announced by Volor

[Rylitah]

Racism and being an antivax covid denier aren't exactly Western specific politics orz, but I do want to say - I'm sure what happened is they heard this guy's songs and contacted him to produce a demo and left it at that without looking deeper into him. I hope this encourages them to look more into the people they commission for work in the future. I am glad for the quick reaction, even if it did happen quietly (which makes sense, considering they seem to have never announced it to begin with).

Speaking of which, thanks for the clarification! I didn't see any hint that Volor had ever tweeted about it to begin with, but it seems clear from the tweet I linked that people found out and informed them at least. It'd be a little funny if he shot himself in the foot by posting about it early - but I don't want to derail this thread much more about this topic, so I'll leave it at this.

 

[peaches2217]

To be fair to Volor, it seems to me that from their tweets, website, and emails, their grasp on the English language is, a work in progress, and I wouldn't expect them to be as familiar with western politics as we are, Like I literally don't know anything about eastern politics. Us warning them and them and letting them know when they've made a faux pas like this is probably the best we can ask for. And their response was quick, the timestamps between those tweets was about 2 hours which has to be a record for a company righting a wrong.

Volor's a company that's got several people on board from and markets to audiences across the globe, not a single person or completely localized company. If they're going to interact globally, they do have a level of responsibility to be at least surface-level aware of controversial topics in their markets, and them not taking such responsibility shouldn't be brushed off so easily. I believe they messed up, plain and simple, but their refusal to host him proves that, in the end, they're taking that responsibility, and for that I applaud them. It was a mistake that's been corrected. That gives me newfound respect for them. Though I'm not sold on Ellie AI, I'll continue to support Volor, and I look forward to supporting them in the future as well!

 

[lIlI]

Volor's poor attention to detail is becoming more frustrating. They have a history of not checking their work: misspelling their own product names, official announcements with basic errors regarding times and dates, poor communication with subsidiaries leading to public controversy, publishing unfinished surveys and broken payment systems: it speaks to a lack of organisation and care. Each major announcement feels like it's followed by a correction and an apology.

While I have no problem with grammatical mistakes from a non-native company, I expect a group of any size to do the basic due diligence of proof reading and test runs. It's become obvious that this isn't happening, which is concerning as a customer: if a company doesn't bother checking whether their posts are correct or their store functions, have they checked their website is secure?

Before these problems felt slightly too petty to post, but now we have an example of it causing a serious mistake. While it wouldn't surprise me if no one at Volor did a background check on the producers they contacted, that speaks to a broader issue with the company.

I sincerely want them to do well, and hold no grudge against any of the employees if this was a decision made in error. But they need to start doing better.

 

[cafenurse]

I had never heard of this producer before so I was going to give volor the benefit of the doubt that they had done some scandals in the past that you had to have been there to know about or something (because that happens to me all the time, where i like an artist until someone brings up something scandalous theyd done that i didnt know about and then i was embarrassed) but like...I clicked on their twitter and its like 50 tweets in a row of that kinda stuff. I'm not sure...how you miss that. Unless they just listened to a few of their songs and was like "yup lets go with this producer", but i can't imagine not even briefly checking their main social medias....

 

[Krathia]

As someone who loves mature voices, I'm glad that she has this mature tone. I don't really care that much about her boxart matching 100% her default tone tbh, as long as she sounds as good as in this demo.

 

[Silver1063]

Another demo premiering in roughly 2 hours!

 

[lIlI]

New demo!



They released a full version of Artpop too.

We seem to be in a flood of new Synthesizer V demos, after the post-crowdfund drought I'm enjoying it!

 

[lIlI]

My initial thoughts after hearing a bunch of usages: she's a whispery, mysterious sounding voice; absolutely gorgeous in ethereal, intimate arrangements. However, anything with a slightly heavier beat makes her sound goofy in her default state, not so much through fault of the voicebank, but the same way that any very soft, delicate performance would sound awkward in the the wrong genre.

The main issue I have is one that continues from her Standard: very harsh, clicky consonants at the start of words. Luckily there's a lot of plugins designed to reduce this from real life singers (usually default to DAWs), so fingers crossed these are effective on her.

Overall, I think she's an excellent choice for the right style of music, but she definitely isn't an all-rounder.

 

[morrysillusion]

ill put this here because its relevant and i can share my opinions! i got eleanor AI early from the pre-order. she is my first full VB in synth v, i have never bought a synth V voice before now lol. and thats kinda why i preordered her- not having any full banks, wanting some kind of high quality (as in, more HW than the lite versions), and having a discount, i didnt mind preordering as much because i wanted something and synth V is far cheaper than other vocal synths right now.

soooo most of my use this morning has been plug and play obviously. i will say hearing her raw voice- im just really into her maturity in tone. i always love the deeper voices and it seems synth v's english voices are kinda leaning in that direction- mostly with AI at least. likely because it captures the more natural sound of the singer i guess? and not the super specific recording style in Standards that are meant to capture a character voice. anyways- she definitely reminds me of what i like with avanna but a bit more clear. being able to put eleanor in a lot range and not have her be muffled it so nice- of course like avanna she has limits in her range and that soft mature voice gets a little lost on the high notes- but not in a terrible way. just kind of leaves her with a bit of a falsetto/head voice, but it does mean she may not be fitting for a song that needs a more powerful sound at high range, even with the tension parameter. still, her similarity in tone o avanna but with the advantages of more clarity already have me loving her a lot as its definitely a kinda voice that i like, even if its limited in places, it definitely didnt stop people from using avanna from a wide range of genres, and i think some of the parameters like tension may have some advantages too.

i may post a quick test cover soon, but for now id say i really like her even if she doesnt hold too much power- but eleanor never really did much anyways so its not a surprise imo. realizing avanna is almost 10 years old its nice to be getting some voices that can be a successor to her sorta sound after all this time, i dont do much vocal synth stuff anymore but i can surely see myself enjoying her voice.

 

[inactive]

The main issue I have is one that continues from her Standard: very harsh, clicky consonants at the start of words. Luckily there's a lot of plugins designed to reduce this from real life singers (usually default to DAWs), so fingers crossed these are effective on her.

In case anyone wants a fix for this problem, it can be done using TDR's Nova plugin, either the paid one or the free one. Here's a little guide for those who may be interested:

Reducing Plosives with TDR Nova GE | Tokyo Dawn Knowledge Base

docs.tokyodawn.net

And here's a link to the free version of Nova:

TDR Nova | Tokyo Dawn Records

A dynamic equalizer in parallel configuration. Precise tone control paired with powerful dynamics processing.

www.tokyodawn.net

Of course, plosives are best reduced during the recording session via good mic placement, pop filters, etc...

 

[inactive]

Now that I've had a chance to mess around with Eleanor Forte AI, the plosives can be controlled in the editor under Note Properties > Timing and Phonemes > Strength. So if you hear a D or T or whatever that's too strong, just turn it down to taste!

 

[inactive]

One observation about Eleanor Forte AI:

I've noticed that the fundamental frequency of her voice is unusually loud. Granted, the fundamental is usually the loudest part of most pitched sounds (but not always, such as the oboe), but for Eleanor Forte AI it seems louder than I would expect.

This got me thinking... Could this be a result of her singing in a talking/crooning style?

As it turns out, when people talk at normal or quiet levels, their voice overtones are rather subdued in comparison to the fundamental. But when shouting, it's not so much the fundamental that gets louder, but the overtones, even to the point where they are considerably louder than the fundamental. (Every voice is different, of course.)

Although I could only find information in relation to speech, I'm going to make the wild assumption that what is true for speech is also true for singing. This would explain why opera singers can sing above the orchestra, as most instruments have decreasing volume for increasing frequencies of overtones, whereas the human voice can generate overtones that are louder than the fundamental. When an opera singer sings loudly, their loudest overtones (which are above the fundamental) poke above the orchestra's upper overtones (which are mostly quietier than the fundamentals). I suppose this is why power metal singers really belt it out: so they can be heard above the guitars.

Returning to Eleanor Forte AI, the more "open" sounding your song, the better she will work. But the more congested things become, the more you have to be careful.

And keep a watchful ear for the low fundamental poking out of your mix!

 

[uncreepy]

I really like Eleanor Forte AI. I tested her with some rock song MIDIs I had laying around (ex: Breaking Benjamin, Chevelle) and she sounded really cute to me. When you download her, there's a text document with parameter suggestions to get her to sound more like her original voice, and I think they are pretty spot-on and use them with minor adjustments. I also tested her with a very fast song and she was able to sing the words understandably (kind of like sing-song rap), unlike the messy/unusable results I'd get with VOCALOID.

I did have issues with her s/d/t consonants being loud, but changing the phonemes to other ones or carrying the consonant into the next note helps soften them. I still have to test out @parallax_fifths suggestions when I get time.

 

[spiderfriend]

(not 100% sure if posts like this are allowed on this thread even if it's inherently related to the topic and if theyre not allowed im so so sorry ;;)

when the eleanor ai lite drops

 

[Eviltreat16]

Not gonna lie - that's what I did. I need to play around with her more but omg I love her.

I played around with the similar voice suggested parameters as well and it was pretty similar. First Vocal Synth Purchase (besides SynthVPro editor) for me and boy am I enjoying her.