Decrease input level of a USB connected microphone to almost zero?

[Marty Markoe - eMicrophones, Inc.]

In the microsoft.public.windows.vista.general section at
http://www.microsoft.com/communities/newsgroups/list/en-us/default.aspx?&lang=&cr=&guid=&sloc=en-us&dg=microsoft.public.windows.vista.general&p=1&tid=91146c44-5101-46ec-8570-558b8359059f
is an interesting discussion concerning anomalies in the Vista audio sound
sub-system as it pertains to WSR (Windows Speech Recognition).

The gist of the thread is that the audio subsystem changes the microphone
recording level to a maximum level no matter what you set it to. This higher
recording level causes WSR to pick up stray sounds it normally would not pick
up. One simple solution has been to lower the recording level to one half of
what is set when you do the Microphone set up. However, annoyingly, the level
is sent back to maximum at some point without your realizing it.

One suggestion from a member of the Microsoft speech workgroup was to
disable AGC or microphone boost. In my case this was not applicable as there
was no setting for boost. Another suggestion was to use the Gold Wave
recording software to set the microphone recording level to 30% and it would
be kept at that point. I hope to test this out later today.

Before leaving the office yesterday I found in the Advanced tab of the
Microphone Properties, settings for Exclusive Mode were enabled for:
1. Allow applications to take exclusive control of this device
2.Give exclusive mode application priority

Has anyone played with these? I will try disabling/enabling the various
combinations and see if that helps and post back here.

Marty Markoe, eMicrophones, Inc.
See us at: http://www.mymsspeech.com

----------------
This post is a suggestion for Microsoft, and Microsoft responds to the
suggestions with the most votes. To vote for this suggestion, click the "I
Agree" button in the message pane. If you do not see the button, follow this
link to open the suggestion in the Microsoft Web-based Newsreader and then
click "I Agree" in the message pane.

http://www.microsoft.com/communities/newsgroups/list/en-us/default.aspx?mid=e72ebc93-1c9c-45c9-ab06-7f16e7d9cf63&dg=microsoft.public.speech.desktop

[Mark Conrad]

In article <E72EBC93-1C9C-45C9-AB06-***@microsoft.com>, Marty
Markoe - eMicrophones, Inc.

Post by Marty Markoe - eMicrophones, Inc.
sub-system as it pertains to WSR (Windows Speech Recognition).
The gist of the thread is that the audio subsystem changes the microphone
recording level to a maximum level no matter what you set it to. This higher
recording level causes WSR to pick up stray sounds it normally would not pick
up. One simple solution has been to lower the recording level to one half of
what is set when you do the Microphone set up. However, annoyingly, the level
is sent back to maximum at some point without your realizing it.

Marty, thanks for tackling this problem, wish you luck.

The problem is keeping me from making optimum use of
one of my microphones. ("theBoom" microphone)

If you sell that mic, I am certain that you would be able to
boost your sales to those people who need to use that mic
in a very noisy environment, provided someone cracks the
problem of reducing the mic sensitivity within Windows,
without appreciably deteriorating the signal-to-noise
ratio of the microphone circuit.

The aim of this exercise is to force a user to place the mic
extremely close to his mouth, say within a half inch,
or even a quarter inch, in order to make his voice as loud as
possible, compared to the background noise, thereby
improving text recognition accuracy in a noisy location.

This works very well with Dragon, because Dragon is better
at ignoring extraneous "breathing" noises, compared to
WSR. (Windows Speech Recognition)


FWIW, the suggestion of that member of the Microsoft speech
workgroup to disable AGC is a bad idea, mainly because AGC is
really needed to "even out" the volume fluctuations of a users
voice, especially when the mic is very close to his mouth.

e.g. "the RAIN in SPAIN falls MAINLY in the plain"

Users unfortunately tend to have wild fluctuations in the
volume of their speech, especially when they get excited
or emotional.

Recently, I was listening to an executive member of NASA
on my TV, that executive had a very soft voice, I had to
strain to hear what he was saying.

However, he had a VERY LOUD laugh, which almost
deafened me whenever he let loose with his laugh at
unexpected intervals, interspersed with his soft talking.

I can just imagine what would happen if WSR was trying to
transcribe his speech to text, with AGC disabled.<g>

Mark-