Digital voices for animation - Lovo

Hello everyone!

I came across this site offering digital voice over for animation, among other things, and I must admit that i am very impressed by the voices quality.

LOVO - AI Voice Generator & Text to Speech

Voice over is a necessity for many Blender productions and I wonder if this is the future of voice-over for indie productions, also, I would like your opinion on the subject!

1 Like

I would love to see some vocal puppetry solutions. Text to speech is impressive, but it is hard to get nuance.

I can do all the voice acting myself (I may not be very good, but I can) , but I can’t do a range of different voices or accents.

It would be nice to record the performance you want, emotionally etc and then change the voice to another character.

I don’t know why this doesn’t exist yet. I feel like the above solution with Lovo is actually a harder task.

I agree, it would be nice to record a session in your own voice, with all the emotions and such, and then just change your voice to a new voice, I have no idea if such a thing exists or is in development, I looked it up but found nothing! :thinking:

As for Lovo, it’s simple enough, you upload a script, choose the voice you want, then you download a wav file, really simple, I will try it for my next tutorial, as it seems easier than recording my own voice over.

From what i understand, they are planning to add control over emotions in the near future, so my guess is that will be done through the script you upload, again, my guess is that they will use tags wrapped around part of the text where you specify a certain level of emotion!?

Right now the quality of Lovo’s voices seems very impressive, and they say they will make it so we can use these voices for animated films, if that’s the case, I think this could be very interesting.

1 Like

The people behind Descript have this figured out. Their tech is able to edit audio as if it were text, and NEW words can be introduced as well. So you record your own voice, and the text is transcribed to text. With the standard version it will allow you to edit the text and remove parts, or add new parts that weren’t there originally.

They have the AI in place to replace the entire audio with a new voice as well, as long as the AI has sufficient sample voice data to work with. And the original intent and expression is left intact!

Now, before you get all excited: the company decided to block the general use of this, and they have to be contacted before they allow anyone to change a voice for another one.

The reason is clear-cut as crystal: if anyone could replace their voice with someone else’s, just imagine the pandemonium by those who abuse this. You could take the voice of any celebrity, and use those voices in your work. What about copyright? Or worse, fake news creators would have a field day with this.

Can you copyright a voice?

As to misuse for fake news, that genie is probably well and truly out of the bottle. I imagine who ever you don’t want to have access to technology like this already has access to technology like this.

The phrase the camera never lies, whole never completely true, is well and truly dead.

So open it up for the rest of us, I want Al Pacino to be a cohost on my podcast

Less facetiously, I’d be happy enough with a bank of generic different voices, based on every one, but no one in particularly

Lovo basically allows that, just look at the videos, there is also LyreBird, a Montreal-based company that allows the creation of any voices from a sample, not sure you could use a different voice than your own, but the technology is certainly there!

LyreBird AI

Lyrebird is text based as well

You are only supposed to use your own voice :wink:, but what’s the point of that, I have that already.

They do have generic voices too, which I have used. The free plan is quite limited, 200 characters of speech. And the most basic paid plan is quite expensive. $50 a month.

However, the technology is here, just not fairly distributed yet