I am downloading Cepstral David Windows 6.2.3.801 now. I will try it out and let you know the results.

Okay, so here is what happened.

I closed ARC. Everything was working fine with voices (just the standard windows voices and being able to change them).

Installed Cepstral David Windows 6.2.3.801

I opened a new project and added the Speech Setting and Speech Synthesis controls.

I set the voice to Microsoft David male and clicked the Say button on the Speech Synthesis control. It sounded normal.

I set the voice to Female with the same voice and it sounded like a female.

Every other voice I choose sounds like Microsoft David, either Male or Female.

Id say there is an issue. I will dig into it more tomorrow to see if there is something else that is causing it. I also have .net developer studio so I will play with it within an application to see if there is something that can be done in code to make the other voices work.

"Only Microsoft-signed voices can be used by SpeechSynthesizer. Here is a list of Microsoft-signed voices provided with Windows."

This is from this site.https://msdn.microsoft.com/en-us/library/windows.media.speechsynthesis.speechsynthesizer.voice.aspx

My assumption is that the voice I installed did something it wasnt supposed to do in the registry. I am still digging.

Okay, so yes, these will not work with the us version of 8.1.

It did work in windows 7 but not in windows 8.1 if you have a US version of the OS. This has to do with .net developer studio and specifically the System.speech.Synthesis.SpeechSynthesizer object that is included with the version of windows you are running. This component has nothing to do with EZ-Robot, but is a limitation of 8.1. They will work fine if someone has written a driver to make them work (which is why they work in other areas outside of .net 4.5.0 or 1). They will work fine on your system but will not work natively with the System.Speech.Synthesis.SpeechSynthesizer object that .net 4.5.x is running off of on specifically Windows 8.1.

I will see if there is another object that can be used but I doubt that DJ will see the need to change this out. You need to find specifically TTS voices and not something that is trying to be converted to TTS.

As to what is happening? My assumption is as follows. In ARC, there is a try catch and finally around the part that loads the voices into the list.

Probably something like this but a drop down list,

try
{
.
.
.
foreach (InstalledVoice voice in ss.GetInstalledVoices())
{
      textBox1.AppendText(voice.VoiceInfo.Name.ToString());
      textBox1.AppendText("~");
}
.
.
.

This part of the code works fine. But when you try to do

SpeechSynthesizer.SelectVoice("Cepstral David");

there is an error, so what happens is it keeps the default name. The error that happens is System.ArgumentException: Cannot set voice. No matching voice is installed or the voice was disabled.

at System.Speech.Synthesis.SpeechSynthesizer.SelectVoice(String name) at EZ_AI.SpeechTest.SpeechTest_Load(Object sender, EventArgs e) in whatever path to the CS file:line whatever line it is

This is odd because the list is being populated with the installed voice names and the user is selecting the name from the list. Unfortunately this doesn't work right in the US..

I am working on how to get it fixed right now because even with an uninstall of the voices, it doesn't fix ARC. I was able to change names in my little test app as long as I didnt select the non TTS voice.

Going to bed now. I will dig in more in the morning.

@David.

Thanks for your response. I really appreciate you looking in to this, and I wish you luck finding a solution. Interesting to see you encountered issues as well. Really curious why it works for some and not for others using the same OS.

@merne.

Thanks for your input my friend. As Alan has already stated the issue is for Windows 8.1, but I mentioned earlier that I have also tried setting all the voices to the system default voice within "Speech recognition" > "Text to Speech" to no avail. Just to say though that I never had any problem with windows 7. In fact, I could leave another TTS voice as system default and Cepstral would still work in ARC. I think the settings were Male Senior.

@Alan.

It's still niggling at me thinking it's something to do with Regonal settings. But I have tried changing the them with no joy, although I havn't tried this with system set up as I thought it amounted to the same thing. Maybe it doesn't. Like you say, most People in Canada do not seem to have an issue, but then there's Rich, in the UK, the using Cereproc for Javis (although not sure what operating system he is using with this). I'm just hoping the guys at the office are in the process of figuring something out, but it certainly looks like David C is on the case.

@Steve.... You got me curious too why it seems to work for me.... I am probably way off (as usual LOL) but Canada has 2 official languages.... That being said maybe that has something to do with why mine works?

@Richard.

Your way way way WAY off. Only joking bud. You may have a point. I didn't know Canada has two official language's. But then in the UK, why does Rich's work and mine doesn't? Same voice, and assuming same OS.

@Steve... Yep English and French are Canada's two official languages.... Not sure you should create another thread but maybe you can ask others to start posting there results in adding voices to use with ARC? I think we need more people to comment so we might be able to see some sort of pattern develop...