[LAU] capturing spoken text -- alsa settings

[Susan Cragin]

I am running Dragon NaturallySpeaking on Linux through WINE. 
One of the benefits to doing so is that you can tinker with the sound settings using ALSA mixer. 
I would like to tweak it to give the maximum amount of sensitivity to the range of the spoken human voice. So far:
front mike = 67
mike = 61
l/r capture = 61
digital = 63 
gives pretty good results. 
Any ideas? Plugins to install? 
Web page with helpful hints?
Thanks