I apologize for starting a thread about this ,I tried to search as it seemed like an issue that would be fairly common,I probably didn't phrase my queries correctly.
Anyway my problem is this,as most folks I like to have the program interact with me through speech output as much as possible and I've been very successful in achieving high accuracy for recognition in my commands however I find that unless I output to my headphones (which I rarely do) the mic picks up VC's replies and of course begins trying to understand and act on them.I know this isn't good for my overall recognition experience as in essence Im allowing another voice to train when it should only be learning mine.
My setup is output through Realtek Digital Optical to my home theatre and I really enjoy it that way.Im using a mid range SteelSeries Headphone set with Mic Boom.As I said I usually just plug in the IN and not the OUT.
I seem to recall hearing James mention audio ducking in one of his videos but I haven't had any luck finding it again.Of course I could just turn the volume down low but that seems rather lame.
Again sorry to clutter up the place with something that maybe an oft' answered question but you'll find I don't post much anyway so this can count as my contrib for October
.
Thanks , and thanks to JitterJames and all the other people who put so much time and energy into expanding this program.It really is wonderful.
JonPeyton
