Warning: The following steps involve editing the registry. Continue at your own risk.
This tutorial is based on Voran's guide found here. The steps he outlines is for Windows 7 64-bit. If you have a 32-bit system, then follow the instructions below. Also, please note that I was only able to make this work with US voices when used with Microsoft Word. For an excellent tutorial on how to enable text-to-speech in Word, watch this video.
Now, on with the steps.
1. Download the SpeechPlatformRuntime.msi file for x86 systems from http://www.microsoft.com/en-us/download/details.aspx?id=27225
2. Just double-click the file and follow the installation instructions.
3. Download the two US voices Helen and ZiraPro from http://www.microsoft.com/en-us/download/details.aspx?id=27224 . As I said earlier, these were the only voices I got to work.
I tried the procedure with GB_Hazel and CA_Heather but no joy. If you get these voices to work, please leave a link or comment below on how you were able to make these work.
4. Again, just double click each .msi installer for each voice and follow the installation instructions.
5. Now it is time to edit the registry. Run regedit. Initially, if you go under Under - HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Speech\Voices
you should only see Microsoft's default voice Anna. While if you look under
HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Speech Server\v11.0\Voices
you should see the two voices you installed (Helen and ZiraPro). If you're wondering why my image has four voices, refer to my comments above.
Our goal is to edit the registry so that the image above becomes like the following image
6. To do that, right click the "Tokens" folder under
HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Speech Server\v11.0\Voices
and choose Export. Save this file to your desktop as voices1.reg so it will be easy to find later.
7. Open voices1.reg in Notepad. Then do a Find-and-replace. Enter \Speech Server\v11.0\ into the "Find What" field and enter \Speech\ into the "Replace With" field. Click "Replace All". Then save the file. Note, after you save the file, its icon might change into a notepad icon, meaning you might have saved it as a text file instead of a registry file. If this happens, make sure you save the file with a .reg extension instead of a .txt. If the file still won't open using the registry editor, try right-clicking it and opening using regedit.
8. Double click the newly saved voices1 registry file. You will then receive a registry warning prompt. Click Yes.
9. To check if you were successful, go to Control Panel > Ease of Access > Speech Recognition > Text to Speech.
10. You should now be able to select additional voices (Helen or ZiraPro) aside from Microsoft Anna in the dropdown box. You can then press on Preview Voice to check if they sound right. If you're satisfied with the voice, then click on OK. Now, you should be able to use the new voice to read portions of your word document. Really useful in proofreading, and hopefully less annoying than Anna.
I hope this tutorial was able to help you out. Feel free to leave me feedback below.
Addendum: I tried changing the voices in Narrator and it was able to recognize GB Hazel and CA Heather. But MS Word still wouldn't recognize these two non-US voices. Again, if you are able to come up with a solution to this issue, please let me know. Thanks.
This tutorial is based on Voran's guide found here. The steps he outlines is for Windows 7 64-bit. If you have a 32-bit system, then follow the instructions below. Also, please note that I was only able to make this work with US voices when used with Microsoft Word. For an excellent tutorial on how to enable text-to-speech in Word, watch this video.
Now, on with the steps.
1. Download the SpeechPlatformRuntime.msi file for x86 systems from http://www.microsoft.com/en-us/download/details.aspx?id=27225
2. Just double-click the file and follow the installation instructions.
3. Download the two US voices Helen and ZiraPro from http://www.microsoft.com/en-us/download/details.aspx?id=27224 . As I said earlier, these were the only voices I got to work.
I tried the procedure with GB_Hazel and CA_Heather but no joy. If you get these voices to work, please leave a link or comment below on how you were able to make these work.
4. Again, just double click each .msi installer for each voice and follow the installation instructions.
5. Now it is time to edit the registry. Run regedit. Initially, if you go under Under - HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Speech\Voices
you should only see Microsoft's default voice Anna. While if you look under
HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Speech Server\v11.0\Voices
you should see the two voices you installed (Helen and ZiraPro). If you're wondering why my image has four voices, refer to my comments above.
Our goal is to edit the registry so that the image above becomes like the following image
6. To do that, right click the "Tokens" folder under
HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Speech Server\v11.0\Voices
and choose Export. Save this file to your desktop as voices1.reg so it will be easy to find later.
7. Open voices1.reg in Notepad. Then do a Find-and-replace. Enter \Speech Server\v11.0\ into the "Find What" field and enter \Speech\ into the "Replace With" field. Click "Replace All". Then save the file. Note, after you save the file, its icon might change into a notepad icon, meaning you might have saved it as a text file instead of a registry file. If this happens, make sure you save the file with a .reg extension instead of a .txt. If the file still won't open using the registry editor, try right-clicking it and opening using regedit.
8. Double click the newly saved voices1 registry file. You will then receive a registry warning prompt. Click Yes.
9. To check if you were successful, go to Control Panel > Ease of Access > Speech Recognition > Text to Speech.
I hope this tutorial was able to help you out. Feel free to leave me feedback below.
Addendum: I tried changing the voices in Narrator and it was able to recognize GB Hazel and CA Heather. But MS Word still wouldn't recognize these two non-US voices. Again, if you are able to come up with a solution to this issue, please let me know. Thanks.