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

I'm using Android 4.4.2 with Google TTS Engine. My Phone Brand is Huawei. I can't find any culture of zh-?? and Chinese voice option buttons like your screenshot. I'm sure I installed both Voice Package of Chinese Mainland and Chinese Taiwan. Another Android Demo did speak Chinese.

 

I tested it under various devices and this is how it looks on my FP2 with Android 5.0:


Sorry, this has something to do with your (old) device and I can't solve this...
But you can try this:

http://www.androidpolice.com/2015/0...w-languages-including-cantonese-and-mandarin/


Cheers
Stefan

 

Yes, thank you, that was very helpful!

 

You're welcome and have a nice day!


Cheers
Stefan

 

Hi,
Thanks for your reply. I've downloaded the latest Google TTS Engine lately cause the original Pico TTS doesn't support Chinese. So I think maybe it is caused by the old Android 4.4. I will try to get an Android 5.0 Phone and test it again.

Best,
Yang

 

Hi Yang

That sounds good - please let me know if it worked.


So long,
Stefan

 

Hi Stefan, yesterday finally we bought RT-Voice asset but payment was done by another person (a partner of mine). How can I download it with my user?
I have the receipt order.
Thanks!
Leo

---Update--

How can I download the purchased RT voice from my Unity User? By mistake was bought with the unity user of my partner.
Please Im in a hurry!

 

Hi Leo

That's something that must be solved by Unity. They have to re-assign the purchase from your partner to your account.
Please file a support request.

Meanwhile, there are two possibilities:

1. Your partner downloads the asset, imports it into an empty project and exports it for you
2. You send me the invoice and I can give you the latest version in return

Both solutions aren't great, so please make sure that Unity can help you further.


Cheers
Stefan

 

I´m going to send you an email right away.
Thanks!!!!

 

I've sent you the package.
If you have an questions or suggestions, just write me an email.

So long,
Stefan

 

Hi Stefan, another question!

In the Asset Store description, RT-Voice is described as being able to:

"Fine-tune your voices with rate, pitch and
volume
• Current word, visemes and phomenes on
Windows and iOS (incl. mark functions) "

Does this mean that it is possible to alter rate, pitch, and volume at run time? Ie, a frightened character might speak faster or quieter, an angry character might speak louder, etc, etc?

How extensible would this system be to modifying the way words are pronounced or expressed at run time?

Thanks!

 

Hello again

Yes, you can modify the rate, pitch and volume for every "speech" (= text).
Take a look at the platform-specific limitations in chapter 2.

About the pronunciation- this is currently not possible due the underlying TTS-systems. The only TTS that supports SSML is Windows (and MaryTTS).
We will try to add this feature in a future release but I can't promise anything right now.

Meanwhile, you can try to add even more variance with Unitys own audio filters (effects like distortion etc.).

Have a nice weekend!


Cheers
Stefan

 

Hi,

I have purchased the standard version of RT-Voice, and I am very happy with it so far. I was able to use it very quickly.

I am currently trying to add some additional voices to my system to use with RT-Voice, and I have downloaded a free trial of these IVONA voices:

http://visionaid.com/phpincludes/en/support/voices/voices.php

However, I can't seem to see them in the Control Panel, or Unity. I suspect voices from Cereproc would be the same, but I don't want to spend money on them until I can solve this issue.

I believe this is because I am using a 64-bit version of the Unity editor, and the voices are 32-bit only:
http://www.webbie.org.uk/blog/microsoft-speech/

Are you aware of a solution to this?

Or, do you know if the voices available from IVONA or Cereproc work with Rt-Voice in 64-bit Unity?

Thanks!

 

Hi

Thank you for buying "RT-Voice"!

Yes, this is already solved. Simply enable "Enforce 32bit Voices" in the "Configuration":



But be aware that you are not allowed to redistribute the voices with you game (except you have a contract with Cereproc)!
Therefor, I would recommend using MaryTTS instead...

Anyway, I hope this helps!


Cheers
Stefan

 

Hi,

That worked great, thank you!

I am aware that I can't distrubute the free voices, I was using them just for testing. Cereproc do offer business licences, but I don't know if they allow distrubution, that's certainly something I will need to look into in the future, MaryTTS may end up being a better option.

MaryTTS is an online service though, isn't it? Which would make it slower, if I'm not mistaken?

 

MaryTTS is open-source and it runs best on your own server or as a local installation on the end-users machine (if you develop for standalone).
The delay is barely noticeable and RT-Voice can pre-generate speeches, so it should be no problem at all.

I hope this is helpful.


Cheers
Stefan


Edit:
You can test it with our demo scenes - just enable MaryTTS

 

That sounds like the best idea then, the Cereproc distribution licences are extremely expensive.

Thanks again!

 

It works offline?

 

Yes

RTV uses the underlying TTS-system on Windows, Mac, Android, iOS and WSA (UWP) => none of them needs an online connection.
For the other platforms we use MaryTTS which can work online or offline.

I hope this helps you further.


Cheers
Stefan

 

Hi Stefan,

good work you guys have done there! We are using your Asset to "voice act" our ingame Avatar which is a robotic drone. The TTS sounds really good and fits our Avatar well - even without a voice actor.

I have a question concerning MaryTTS. I want to ensure that RTV uses the same voice on all devices. Since we are building for Deskop devices only, I am fine with the additional build size. Can you explain how to include a local MaryTTS in a Unity build?

Thanks in advance!

 

Hi

That's easy, just follow this tutorial:

http://fastertutorials.com/2013/08/13/35/

This MaryTTS-directory must be part of your installation application.
Also make sure that you start the server when needed (e.g. as Windows Task or with C# Process-class).

That's all.
We will probably add a bundled version in the future.


So long,
Stefan

 

Hi, Could you tell me what the latest is on getting RT-Voice working for a WebGL build?

The blurb on the asset store and in the docs mentions that it's still in beta. Do you have an idea of when it might be functional?

In the meanwhile could you explain what's meant by needing CORS or policy server.

Many thanks,

Jerome

 

Hi Jerome

WebGL and MaryTTS should work as long as you manage to install a (HTTPS-)server as proxy.
Requests from WebGL to another instance are only "secure" it they are HTTPS and/or have "Access-Control-Allow-Origin: *" in the response header.

Unfortunately, we didn't have enough time until now to tinker a solution together. But in theory, it should be not too hard...

Here some details:
https://en.wikipedia.org/wiki/Cross-origin_resource_sharing
https://enable-cors.org/server.html
http://www.dfki.de/pipermail/mary-users/2014-March/001522.html
https://github.com/marytts/marytts/wiki/Local-MaryTTS-Server-Installation

I hope this helps you further. If you found a solution, please let me know


So long,
Stefan

 

Thanks for the info to get me started. I'll start looking into it and report back

 

Thank you and have a nice weekend!


Cheers
Stefan

 

Hi Stefan,

I'm thinking of buying your asset to use it on a vr ps4 proyect. Do you know it will work on ps4?

Thanks in advance.

Manu

 

Hi Manu

Thank you for your interest in "RT-Voice"!

We added MaryTTS as an independent voice provider for ALL platforms and it should work like a charm.
But: you have to have an internet connection available to use MaryTTS.

I hope this helps you further.

Have a nice weekend!


Cheers
Stefan

 

Hi Stefan,

Thanks for your quick response. Is there any possibility that your asset work without internet connection?

Thanks
Manu

 

Hello again

Unfortunately, we don't own a PS4-devkit, so we could never test if there is ANY native TTS-system on the PS4...
I just found this and it looks like there is only an English-voice available: http://forum.audiogames.net/viewtopic.php?id=20807

So the short answer is no
But we're convinced that most of the players have an internet connection nowadays - so the poor souls without an Internet connection won't hear the speeches (but your game will just work fine - RT-Voice wont't crash or something if it can't speak).
There are many possibilities to reduce the impact of that issue (like texts and pre-generated "fallback" audio speeches etc.).

So long,
Stefan

 

Last Chance? is the sale over tomorrow?

 

There are some rumors

 

Hi,

I just upgraded my project from Unity 5.5 to 5.6, and I am having now getting a single error when using Rt-Voice.

Calling Speaker.Speak now results in:

"MissingMethodException: Method Not Found: 'UnityEngine.WWW.GetAudioClip'"

I am using Rt-Voice 2.7.1, Windows 10 64-bit, fully updated (Creators Update). The "preview voice" option in the inspector works fine, and all of my Add-In voices are displayed properly. The code also works fine in 5.5, and there are no other console errors or warnings.

From reading through this thread, it seems that another 5.6 user (on a mac I believe) previous encountered this, and it was fixed in a later release of Rt-Voice. I'm not sure why it's cropping up again, all of my software is up to date.

Thanks a lot for any advice.

 

You have to delete "Assets/crosstales/RTVoice" and re-download the package from the AssetStore within Unity 5.6 and re-import it.


Cheers
Stefan

 

That's great, that worked perfectly, thank you!

 

[EDIT] I needed to reboot the laptop and things fixed themselves Leaving this here in case someone else runs into the problem...

After upgrading to the latest RT Voice Pro by first deleting the folder, then DL/Import, I am continuously getting the following error (MacOS) when running any of the demo scenes. The speech comes through with a default voice but the program always starts with this red console error and brief hang in the editor. Any thoughts?

Could not get any voices!
System.InvalidOperationException: The process must exit before getting the requested information.
at System.Diagnostics.Process.get_ExitCode () [0x00000] in <filename unknown>:0
at (wrapper remoting-invoke-with-check) System.Diagnostics.Process:get_ExitCode ()
at Crosstales.RTVoice.Provider.VoiceProviderMacOS.getVoices () [0x00000] in <filename unknown>:0
UnityEngine.Debug:LogError(Object)
Crosstales.RTVoice.Provider.VoiceProviderMacOS:getVoices()
Crosstales.RTVoice.Provider.VoiceProviderMacOS:get_Voices()
Crosstales.RTVoice.Speaker:get_Voices()
Crosstales.RTVoice.Demo.GUIMain:Start() (at Assets/crosstales/RTVoice/Demo/Scripts/GUIMain.cs:45)

 

I'm glad everything works now!

So long,
Stefan

 

Hi,
I am using rt-voice dll to develop the unity project but after building it to windows platform .We are facing some issues related to dll's. Please review the screenshot attached and suggest any solutions..

For your information, we have used Unity 5.5.3 for creating project. We see this issue in both 5.5.3 as well as 5.6.0f3.

We are able to execute the project in Unity and able to see the result but facing issue only after building it for Windows platform and trying to execute using Visual studio 2017.



Thanks
ravinder

 

Hi ravinder

We never heard of a problem like this before.

Why do you suspect RT-Voice to be the problem? Depending on your screenshot it looks like the AssemblyConverter failed and our asset is only the last dll in a bunch of other dlls...

Which RT-Voice-version and OS are you using?


So long,
Stefan

 

P.S.: for 5.6 you have to re-download RT-Voice from the AssetStore - Unity changed some basic stuff and therefor, we had to create two different versions for the store.

 

Hi Stefan,

Thanks for your response.

I am using RT-Voice Pro - 2.7.1, Unity 5.6.0f3 and Windows 10 x64 bit OS

Thanks
Ravi

 

Hi Stefan,

I tried to re import the RT-Voice 2.7.1 from Asset Store, even then i see issues when i build and run the project on Unity. Please find the screenshot below for logs.


Thanks
Ravi

 

Hi Ravi

Which is your target platform: Windows Standalone or WSA (UWP)?
For the WSA-platform, you need the source code to build - see chapter 9 of the documentation.

I hope this helps, please let me know.


Cheers
Stefan

 

Hi Stefan ,
1. I am using UWP platform and build the project ,
2. Open with visual studio Enterprice 2017 but still i am facing same issues in visual studio.


Thanks
Ravi

 

Did you extract the sources?

 

Hi Stefan ,

I follow the steps mentioned in chapter 9 in the documentation, but after build the windows project i faced same issues in unity and visual studio projects , I already posted the issues (Screen shots) yesterday.

Note:
I am using below versions:
1. Unity 5.6.0f3
2. Windows 10
3. Visual Studio Enterprise 2017.

What do you mean "Did you extract the sources?" in your last reply i can not understand this point.

Please let me know.

Thanks
Ravi

 

It looks like you didn't delete the "RTVoice.dll". Please follow exactly those steps:

1. Create an empty scene inside your project
2. Delete the "Assets/crosstales/RTVoice/Plugins/RTVoice.dll"
3. Import the "Assets/crosstales/RTVoice/Sources.unitypackage

If you still encounter an error, it should not mention "RTVoice.dll"...

 

Hi Stefan ,
I did follow above 3 steps exactly and build the windows project i see few issues in unity and windows project not created due to this errors.
Here is the screenshot:



can you please check once and let me know the information .

Thanks
Ravi

 

Ok, we found the problem. Please send me an email with the invoice and I will send you our latest build.

 

Thanks for your response will send you details soon..

Regards ,
Ravi

 

I've just sent you an email with the fixed version.
Please let me know if it worked.


So long,
Stefan