Training with selected documents

rreilly wrote: I ran the training with the text provided and it seemed to work OK.  Then I pasted my own document in and when I click on train.  It gives me this message: "Speech Training is not configured properly.  Please go the the Speech Recognition Control Panel, click on Advanced Options and set Microsoft Speech Recognizer 8.0 for your system language as the default recognizer."  When I go to the control panel and advanced options it tells me that the default is already set to Microsoft Speech Recognizer 8.0.   I cannot seem to solve this problem.  Can anyone help?

We assume when you say you ran the training with text provided, you were using the Accuracy tab feature? If so, this feature is meant to relate how you pronounce phonemes (individual sounds). As The Rainbow Passage has a representative distribution of the 44 phonemes in the English language, more training in this feature should not be necessary.

The feature you should use for adding your typical documents is the Scan feature. This improves your user profile by comparing the words before and after each word for Context.

Marty

I understand how the scan feature works but I went back and looked at the accuracy section.  It presents the rainbow passage but also says, "It helps if you add on or two of your personal typical documents and read those.  Simply copy those documents and paste in the Text Field to the left".

Update to rreilly. My programmer and I went through the Accuracy tab extensively. What we found out is that there must be a paragraph break after each line like you see in, "The Rainbow Passage." This is apparently is something we cannot control and it is how the acoustic performance enhancer from Microsoft is designed. If you wish to add any of your own documents, although I do not feel it is really necessary, paste into the Accuracy text field go to the end of each line and hit the return/and enter key. You should be able to read the text and improve your acoustical model at that point.

Marty