[APP][Pro] Google Services - App for Translate and new Text-to-Speech for any device with url input

Apps
,
505

Do you mean for translate or TTS?

And which JSONs did you edit?

506

I meant translate… checking it again, seems that time I took a copy of your repo, I duplicated actually 1.9.51, while I see your actually last prod. version got 1.10.2 :wink: So no need right, as you changed the method in the meantime.

507

Hello,

I am looking for a solution to have something said locally from Homey from a speaker. What exactly is text-to-speech (non-cloud)? Does this work locally, between Homey and the Google specker?

508

Well,

The TTS app lets you generate a MP3 file through a Google Service, containing speech.

That MP3 is stored on Homey and available through a URL, so you can play it on devices in your network.

The non-cloud version still uses an online service of Google, but you do not need to create a Google Cloud Account for it to work. It is very limited though, the Cloud version is way way better.

509

Okay Thanks :see_no_evil_monkey:

510

I’m not getting prompted for “Project ID” and “Service Account JSON” during the setup/when adding device and there’s no such fields in the device settings? I am adding the correct device , “Text-to-Speech”, not the “Text-to-Speech (non-cloud)”?

Thanks

511

Goto the app settings :wink:

512

Well, that’s embarrassing, I was looking at the “device” settings, completely forgot about the actual app settings which I found in the meantime. However, I’m still getting the “Did you configure the JSON and Voice in the settingspage?” error, although I copy/pasted the JSON that was downloaded from the Google Cloud and selected a voice from the dropdown (I also entered the Project ID). Not sure what I’m doing wrong.

Thanks.

513

I have installed the app, created a Project, added a service user, downloaded the JSON and pasted it into the settings. I have also filled in “Project ID”. But the list of voices are empty and the action card “start saying…” is failing.

EDIT: The instructions told me to enable TTS and then select/create project. TTS had to be enabled INSIDE the new project. Then it worked.

Question below is still something I do not understand:

Also - when using this combination of two flows, one for “start speaking…“ and one “URL Triggered”, how to differentiate between differend speakers? How dows the “Start speaking” flow tell the several URL triggered flowd wich one to fire and send URL to the correct speaker?

514

Basically: create a TTS device for each separate speaker.

515

Ah, Of course. Brainfart.

Is there a equaly simple solution to that my Sonos speakers seem to loop and play the URL over and over again? “Repeat” on the device is turned off! :slight_smile:

516

Since a couple of weeks one of my home minis stopped working with TTS non-cloud, cloud is working just fine. Do you have any tips to fix this? Already tried the default stuff from google.

517

Well, not through my apps. The Google app only creates an URL/mp3 file to play.

How Sonos handles it has to do with the sonos app or speaker.

But my sonos speakers don’t have this issue, the don’t keep repeating. Is it S1 or S2?

518

reviving an old thread here since TTS isn’t free anymore, perhaps that should be updated. :slight_smile:
It’s now 60 minutes per month and then 0.016c per minute.

519

@Thyg0d Where did you get that information?

I just checked Google Cloud, and it still says the first 4 mil characters are for free.

1000332430
10003324301080×2122 165 KB

520

My TTS stopped working because I didn’t set up billing that’s when I started to check around since adding my CC to a service I have little to no control over and is billed afterwards scares me. :smiley:
Then I stumbled upon this.

And apparently AIStudio also costs so you have to prepay.

The welcome credit is of course gone after 90 days as well meaning you sort of have to pay to use..
Which is too bad because it was really well done by you!!

521

Your link is about Speech-To-Text, not Text-To-Speech, which is, afaik (as seen in the link just posted) still free.

522

Sorry, my bad but it seems to depend on what you choose when enabling the API
Anything Gemini seems to cost?

image
image663×286 20.2 KB

image
image820×269 25.2 KB

Or am I misunderstanding something?
I’m used to MS f*cked up costmanagement but Google was actually even less understandable.. :smiley:

523

Yes true, but I am not using Gemini in this app for TTS. Just regular and WaveNet etc.

524

Aaaah now I get it.. :smiley:

So the choices was not what I chose.. It’s just the pricing, then what is used is what is costing..
And since you use regular and WaveNet then it’s free..

Then I stand corrected. :slight_smile:
Thanks for enlightening me!