Rex: the voice of Gladys

Configuration
,
1

Hello,

Since the release of the feature to speak through Sonos speakers, I’ve been able to cut a lot of links tied to Node-RED. It’s really cool!
Now, live, it notifies me to change the filters on my devices or reminds me to take my medication. It’s perfect!

I just have one piece of feedback, and it’s not a small one. We don’t like the voice haha.
Wouldn’t it be possible to have a choice among multiple voices? Because every time my partner hears it speak I get: « you really need to change its voice ».

So, before switching back to my other TTS system, I wanted to know if there’s any possibility to have a choice or a vote on which voice to use?

I know it’s all a matter of taste and color, but for us it’s our daily life ^^’

Are we the only ones in this situation?

Thanks

cce66 2

:slight_smile:

1 Like
3

I’m not really in favor of that kind of speech.
Especially since I agree with her.

4

I agree that would be a really significant plus! :+1:

We often tend to personalize things that speak to us (Waze/Maps in particular), because it’s a criterion that can be a dealbreaker.

For me, it deserves a feature request!

5

It was a joke @spenceur, it was a joke,

6

Thank you for your feedback @spenceur :slight_smile:

Ah, too bad you don’t like the voice ^^

Can you try some tests on this interface and tell me if there’s one you like better?

7

I was able to test 5 of them but:

Thanks for trying out our speech synthesis! You have reached the limit of available requests on the landing page. You can continue, for free, by signing-up to the platform.

I’ll continue later in private browsing ^^

For now we preferred george ^^’

8

@pierre-gilles, I don’t get any sound (voice) from my Sonos anymore.
When I go through Node-RED there are no issues but nothing anymore through Gladys Plus.
Did you disable it?

9

Personally, I don’t have that voice, but when I listen to the ones on the site, I think Matilda sounds pretty good.

10

For info, during playback, on the Sonos app:

1000016814
10000168141080×2316 136 KB

And as a result the music doesn’t resume.

And no logs on the Gladys side

11

Same, indeed

12

And a female voice instead? ^^ For Gladys it’s better if it’s a woman’s voice anyway :smiley:

You’ve used up all the credits :joy:

13

I’m clearly for the « Grace » voice myself after having tried pretty much everything. That said, even though I’m in favor of a female voice to match Gladys, I also think that, if technically possible, offering a choice of two or three female/male voices would be a real asset.
We have the chance to be able to (if technically possible) allow a bit of flexibility and openness, and we’re clearly in the realm of personal tastes and preferences here — plus the option to choose a pair. The risk is that some people won’t use it for that simple reason.

@pierre-gilles … you are the master of the keys ^^

14

I’m in favor of a male or female voice (I’ll check the Grace voice tonight ^^)

15

While we’re in beta for this feature, I prefer to stick to a very simple setup with a single voice :slight_smile:

We can change the current voice to a better one if you find one!

16

Grace works for me — which voice is it currently? I’ll confirm with madam ^^

17

Currently it’s Serena :slight_smile: (model « eleven_multilingual_v1 »)

18

FYI I switched to « Grace » on the « v2 » model! :slight_smile:

I’d welcome your feedback, I hope this improves the voice quality!

19

Hello,

I’m taking the liberty of bringing the topic up again. The chosen voice is good and the feature very useful in our daily lives.

That said, I find its speed is too fast, which makes the messages a bit directive / dry. Would it be possible to reduce the speed? (I checked on the ElevenLabs website, it’s possible via a slider)

Thanks for your feedback :slight_smile:

20

Hi @PhilippeMA,

Do you have an example of a text that you find too fast? So I can run some tests