Hello!
I found a solution which satisfies my needs. As you pointed out, the speech recognition engine can be created without problems by self-coded programs using the SDK/.NET. That’s what I’m mainly interested in.
So far, everything works flawlessly, except when starting the Speech Recognition UI. However, for me, this is not important.
@commodore: I’m running an English version, with an additional German language pack. I can successfully switch to either language with UI and speech recognition. R2 doesn’t seem to make any difference.
Thanks a lot for your help and efforts! It saved me really a lot of time of experimenting (eg. regsvr32 -> less CLSID copying).
Unfortunately, I have no ultimate solution for anybody who wants to use the recognition UI.
I found no way to solve the “language configuration” error message.
I found some extra info, though:
The above procedure only works for applications running in x64 mode. If you want to utilize speech recognition in applications running in x32 mode, you need to copy a few files/entries from a Vista/7 x64. Otherwise you would receive a ComException like this:
Creating an instance of the COM component with CLSID {} from the IClassFactory failed due to the following error: 80004005.
The files/entries to copy are the following:
- %windir%SysWOW64Speech
- HKLMSoftwareWow6432NodeClasses (Here again, all references with “Speech” or above mentioned *.dlls of the speech recognition)
- HKLMSoftwareWow6432NodeMicrosoftSpeech
Thanks again – my problem is solved!
– asbtime
