RT-Voice - Run-time text-to-speech solution

Thanks! Alas, I get the same error. Below is the full error (two of them, actually) from the console. Also, when the file just finishes loading a message flashes onscreen for a split second--something about "No OS voices found".

 

The same version here, but it did not work. However, with the unpack it is working perfectly now. Thank you.

 

Hallo Stefan, just to tell you that RT-Voice is working with no problems in the new Unity 2019.3.13. I don't know what was going on with Unity 2019.3.12, maybe a wrong installation, although everything else was working well in that version. Regards.

 

Hi !
I am a PlayMaker user, and I absolutely don't understand how to make RTVoice work with PlayMaker. I must admit it is the first time, I bought or downloaded more than 200 assets, many of them using PM actions... I searched in this forum, on Youtube, inside your documentation... Could you provide a working example ? Is there any "very beginner" (not in PM, but in RTVoice tutorial somewhere ?
Hope you can help !

 

Hi

Have you tried the PM examples in the "Assets/Plugins/crosstales/RTVoice/3rd party"-folder?


Cheers
Stefan

 

AHHHHHHHHH didn't see the example scenes ! Thanks ! Sorry !

 

Hi

We are using RT-Voice Pro with Windows OS, sometimes Windows 10 and sometimes Windows 7.

I would like to ask if it is possible to know the voice selected as default in the OS by the user. I can see that the Speaker has a 'DefaultVoiceName' member, but it seems to be always the first voice encountered in the system or the first element of the Speaker.Voices list, "Microsoft David Desktop" in my case. Thanks in advance.

Regards
Aaron

 

Hi everyone. I get voices when I use RT Voice Pro-MaryTTS inside of Unity, but not when I build a standalone Windows 10 game. Is that normal? If not, is there a way to enable MaryTTS functionality for standalone games in Windows? I previously used the Windows 10 default text to speech voices with RT Voice, and they worked in the standalone build, but unfortunately they are limited in number and type. Thanks for your input.

 

I tried the WebGL demo again on Mac Safari, now I don't get any console errors, but there are big red words on the canvas:

No OS Voices found -- TTS Not Possible!

Same thing on Mac Chrome and Mac Firefox. Is that expected? I checked System Prefs > Accessibility > Speech and I do have several voices that work.

Hmm....now I'm getting the same thing on Windows Firefox and Chrome and Edge. So it doesn't work on any broswer...

 

Hello again

Sorry, we did some maintenance on our MaryTTS test-server and it therefore didn't work.
Please try it again


Cheers
Stefan

 

Hi Aaron

To use MaryTTS in builds, you should setup your own server (instructions are in the "Documentation" folder).
However, we issue accounts on our test-server for the development phase of our customers projects. If you like an account, send us your invoice and the desired username via email.


Cheers
Stefan

 

Hi Aaron

Windows has different voices per language and system which makes it hard to determine the "default".
We decided that the default under Windows is "David"; however, you can change the property "DefaultVoiceName" in "VoiceProviderWindows.cs" to an empty string or the voice you like.

I hope this helps you further.


Cheers
Stefan

 

Stefan, Thank you. You had previously provided me with a username and password. I was using the demo username and password not realizing it would make a difference for purposes of the build. I made the simple change of entering the info and it now works in a standalone build. I appreciate the help.

 

So now it works on Mac Chrome...but it takes about a minute for the voice to start up. Not really usable. Is that a known issue?

And I still get this error on Safari (and no voices):

An error occurred running the Unity content on this page. See your browser JavaScript console for more info. The error was:
NotSupportedError: The operation is not supported.

 

Hmm, I just did a test with Firefox&Chrome and it works very well. It takes less than 2 seconds to generate the audio...
Do you have a good Internet connection?

About Safari: I don't know why this is happening; we will investigate this further.

However, have you tried the "WebGL_Demo" (that's the one when you click on the top left button)?

 

Ah! No, i didn't notice that. Okay, this works great...mostly. On Safari, I get voices on the first several voice-buttons I click, but then it stops working. Maybe that's just a quirk of this demo and wouldn't be a problem in a real project? No problems on other Mac browsers.

 

As I said, we have to investigate this; but I blame the combination with (an old) Safari and Unity

 

I see two different instructions for upgrading:
1) In the Readme for our version (Pro 2019.3.1) it says

• 3. Delete the "Assets/Plugins/crosstales" folder from the Project-view

but the pdf manual says

• 1.Delete the "Assets/Plugins/crosstales/RTVoice" folder from the Project-view

Which is it?

Regardless, even when we follow the instructions, our classes lose all the references to the RT-Voice classes.
In Rider we use the Context Actions to fix the reference and it compiles with no error. We switch back to Unity and the fix is somehow undone and the references are lost.

 

Hi

It is the one from the pdf; it's fixed in the upcoming version.
Thank you for pointing this out!

However, you should not lose the references; we never changed our meta-files for the prefabs.
Which Unity version are you using? And what's the RTV-version you want to upgrade to?


Cheers
Stefan

 

I been looking for something like this...I been making a game to home school the kids and been making it, but I needed a good tool to handle Text to Speech. ( I thought I seen this before, but was not 100%. ) I will pick this up. ! Looks like it can address my home schooling issue. so I can finish this game for them...

Nice work!

(PS) Teaching my 5 and 6 year old how to spell, and hopefully program as well in the future.

 

Hi @Stefan-Laubenberger after updating to the latest version I now have two errors in VoiceProviderAzure.cs:
Assets\Plugins\crosstales\RTVoice\3rd party\Azure\Scripts\VoiceProviderAzure.cs(273,83): error CS1061: 'Stream' does not contain a definition for 'CTReadFully' and no accessible extension method 'CTReadFully' accepting a first argument of type 'Stream' could be found (are you missing a using directive or an assembly reference?)
on lines 273 and 435. I'm on Unity 2019.3.14 and using API level 4.x

 

Unfortunately, we made a mistake in the last build which causes the errors you experience.
Please replace the file located under "Assets/Plugins/crosstales/Common/Scripts" with the attachment.
We will release a fix asap.

Please let me know if it helps.


So long,
Stefan

 

@Stefan-Laubenberger It's working for me in 2019.3.13f1 with the updated script, cheers.

 

Yes, works for me as well. Thanks for the prompt reply.

 

Hi Stefan,

which voices can i use on iOS? Is it still only Siri?

Thanks,
Ben

 

Hi Ben

No, RT-Voice had always access to much more voices; please see chapter 17.1.4 of the documentation:
https://www.crosstales.com/media/data/assets/rtvoice/RTVoice-doc.pdf

I hope this answers your question.


Cheers
Stefan

 

Thanks, ok that a few
Found this from 2017 when I was searching for, thats why I asked

 

Ups, this is a really old post... We already edited back in 2017 to represent the actual status:

Anyway, I updated the post to make it clearer.

Thank you for mentioning this and have a nice day!


Cheers
Stefan

 

Additional question, are the iOS Voices generated on device or does it need connection and cause traffic?

 

The voices are local, so no additional network traffic

 

Perfect, thanks!

 

Hello, everyone.
Today I have a problem with MaryTTS on RT voice pro.
Because it is not working now. It asks me to setup my own server at http://mary.dfki.de. I don't know how to do it.
Is there any good man who can help me?
Thanks Alexay

 

Currently, I installed RT Voice Pro 2020.1.2 in my Unity 2019.2

 

Hello, Stefan.
I can't use MaryTTS service with RT voice pro now.
It says I have to setup my own custom server at http://mary.dfki.de. ("TTS with current setting is not possible")
I don't know how tot setup the server in my side.
Thanks Alexay

 

Hi Alexay

There is a manual to setup your server inside the "Documentation"-folder.
We also provide accounts for our test-server which can be used during the development phase of your projects.
Just send me the desired username and the invoice.


So long
Stefan

 

I watched a few videos about this system

can i add voices that are not like a robot

 

Sure, you can use AWS Polly, Azure or Google Cloud Speech, which all sound awesome.

 

Is there a way to change that default during runtime to randomize the voice or to allow the user to choose the voice? I tried to see if I could change VoiceProviderWindows.cs to add a set value to allow me to do that during runtime, but the abstract method does not allow a set value.

I could set the voice to a global variable that could be applied to my scripts, but I think this should be something that could/should be applied on the RT codebase. It would be even better if RT could randomize itself at start if the user chooses, but I'm happy to do that randomization and let the user choose the voice as long as I can set it to the Speaker.

 

Hi

We don't recommend changing the voice providers for that; use "Speaker" instead.
I would do it like that:

1. Wait for the "OnVoicesReady"-callback
2. Then get a list of voices with "Speaker.VoicesForCulture"
3. Randomize the result with our extension method "CTShuffle"
4. You're ready to go

You can find more details in our API:
https://www.crosstales.com/media/da...ass_crosstales_1_1_r_t_voice_1_1_speaker.html

I hope this clarifies the situation.


Cheers
Stefan

 

Thanks for this, Stefan. I'll give it a try and will come back if I have more questions.

 

There's something weird with the asmdef files. We're upgrading from RTVoice 2019.3.1 to 2020.2.2 on Unity Editor 2018.4.1. If I do not include crosstales.asmdef in the import and revert the changes to rtvoice.asmdef (which refer to plugins and rtvoice_common) everything compiles and works correctly. This running on my Mac and Android.

When I bring that same git commit over to Windows, it's still not right. It can't find any reference to NAudio (I've seen this before) and the About pane in the RTVoice tool says the old version.

On Unity Cloud build we get a completely different error, for example:

Code (CSharp):

1. [Unity] Assets/Plugins/crosstales/RTVoice/Scripts/Tool/AudioFileGenerator.cs(95,14): error CS0103: The name 'BitsPerSample' does not exist in the current context

 

We decided to delete the asmdef files in the upcoming version since they create too much hassle for our customers.
Please delete it too, thank you!

 

Hi.
I face very strange behavior whenever I open the RT- voice pro example scene then unity not respond.

 

Hi

That shouldn't happen in the first place.

Anyway, how long does the freeze persist? Does it work afterwards (e.g. in the "play"-mode)?
Are you using any anti-virus software despite the standard engine from Microsoft?


So long,
Stefan

 

No I not using any antivirus.
I need to close the unity because after the situation unity not respond.

 

I don't know why this is happening

But it looks like your Unity version is 2019.3.0, which is pretty old. Especially the .0-versions are well known to cause problems.
Is it possible for you to use a newer version of the 2019.3-line or even switch to 2019.4?

 

Now I testing on Unity 2020 best same result.
if click on any scene then unity respond.
If I want to create new prefab or drag and drop prefab into new scene the same result.

 

What's the version of RT-Voice?
Are our demo builds working?

 

I using Latest version 2020.2.2.
Unity not responding even I try to build.

 

That's really strange and we never heard of a problem like this.
Please contact us via email for a faster and better support.


Thank you and so long,
Stefan