Registering and logging in removes this ad.
Registering and logging in removes this ad.
Cure for a hypersensitive microphone
Hello there,
I'm using the Andrea NC-95. I have an unusual problem. My microphone is unusually sensitive. It hears words like "of" or "the" or "and" when I am not dictating anything, when there is absolute silence in the room. Has anyone had this kind of problem? I can't find an option in Dragon to cure it. I've tried redoing my audio settings, but it seems to do no good.
Also, perhaps a related problem: I have an application (Eclipse) which, while Dragon is on, seems to keep going to the "Run" menu every 30 seconds or so. And then after a few more seconds of my doing absolutely nothing, it will select the "Run..." option on that run menu. As far as I know, no voice commands is being issued whatsoever, since Dragon's little popup shows nothing. Does this sort of thing sound familiar? Is there any way of turning on a Dragon debugging mode to figure out what is going on?
Thanks in advance for your time and consideration.


How do you position your
How do you position your microphone?
Bruce
Try moving the microphone
Try moving the microphone out of the path of your breathing, and the problem will likely vanish.
If this doesn't help, make a sound recording of silence, and try to figure out what that noise is that DNS is interpreting as spoken words.
Scott
That's what is interesting.
That's what is interesting. The Dragon mic bar isn't "going into the green" when it thinks it is hearing something and yet small words dribble out. The microphone is about an inch to the right of the right corner of my mouth and a half inch from my face. I test how much breath is being captured by breathing heavy and looking at the microphone bar. I can breathe hard without any pulse at all registering on the microphone bar.
Agree its not a positioning
Agree its not a positioning problem. What kind of soundcard are you using? How long has this problem been observed? What kind of computer are you using? Where are you dictating -- anything in your environment where the cord might be picking up electrical interference?
Try wiggling the various parts of the cord to see if there is electrical interference -- for example, hold it a foot or so away from where it attaches to the soundcard and wiggle that part, then a foot away from where it runs into the headset, etc.
Also Marty's site www.emicrophones.com has a FAQ telling how to make a recording with Windoze Sound Recorder to hear the overall quality of sound -- running that test would be useful.
Finally, do you observe the Eclipse behavior when DNS is not running? What about when DNS is running but the mike is off? What about when DNS is in sleeping mode?
Bruce
PS: Any other unexplained phenomena in your dictating environment -- spirits, hauntings, telekinetic anomalies, etc.
When the small words start
When the small words start "dribbling" press the minus key in the numeric key pad and turn up your sound and click playback and see what you hear. Do you ever see red?
Dragon Reseller since 1989
Go into the accuracy center
Go into the accuracy center and adjust your audio settings and when doing the first test talk louder than normal when reading. Do you ever see the VU meter go red?
tom
Dragon Reseller since 1989
First of all, I'd like to
First of all, I'd like to thank everyone for the help and comments.
I'm using one of those Creative Labs Sound Blaser Live! 24 bit sound card, an Andrea mic with the battery-powere signal strengthening box (forget the exact name).
I will try the cord shaking and recording tests tomorrow, when I am at the office again.
I do not observe the strange Eclipse behavior either when DNS is not running or when the dns mic icon is sleeping or paused.
I actually tried speaking loudly into audio settings. That did no good. But I did not yell so loud that the bar went red. I will try that again.
Can't type much more. Hands ache. (There is a reason I'm using Dragon!)
Yrs, Ara
ara_vartanian wrote:I'm
I'm using one of those Creative Labs Sound Blaser Live! 24 bit sound card, an Andrea mic with the battery-powere signal strengthening box (forget the exact name).
Is the purpose of the "signal strenghtening box" to power the microphone or to boost the microphone audio going to the sound card or does it do both?
is it broken? my take,
is it broken? my take, Quintijn
SFAIK, Tom Westheimer is the
SFAIK, Tom Westheimer is the most "pro" of people offering to help you, so I would give his suggestions priority.
Bruce
BruceCyr wrote:SFAIK, Tom
SFAIK, Tom Westheimer is the most "pro" of people offering to help you, so I would give his suggestions priority.
Bruce
What am I. Swiss Cheese!!!?
Just kidding.
Chuck
Well now that you've jumped
Well now that you've jumped in, its time to get to work! Tom is on the board with a suggestion, yours is no doubt coming, so I think I'll take a break and watch the pros work
Art isn't going to know what to do with his new wealth of consultative expertise.
Bruce
Okay, I have something.
Okay, I have something. First of all, I noticed that my yellow bar is almost always fluctuating up and down, even if there is absolute silence in the room (even when I'm holding my breath -- my breath seems to make no difference). It is usually around half way to the distance at which point it would be a green bar. For those of you who have the optimal arrangement, when there is silence all around, is the yellow bar nowhere to be seen? Second, shaking the wire at any point causes the bar to land firmly into the green. However, the closer to the microphone itself that I jangle the cord (as opposed to the closer towards the computer), the greater the disturbance Is this supposed to be happening? When I play that back, I hear a giant crashing sound quite disproportionate to the meager sound of my stirring the cord. Might this be a source of the problem? Or might this be another symptom of the hypersensitivity?
Thanks again for all your help.
Oh yes, the battery box is the standard one you need to make the NC-95 work. As far as I can tell, it's not making the signal any louder. Or at least it's not supposed to be doing that.
ara_vartanian wrote:when
when there is silence all around, is the yellow bar nowhere to be seen? Second, shaking the wire at any point causes the bar to land firmly into the green. However, the closer to the microphone itself that I jangle the cord (as opposed to the closer towards the computer), the greater the disturbance Is this supposed to be happening? When I play that back, I hear a giant crashing sound quite disproportionate to the meager sound of my stirring the cord. Might this be a source of the problem? Or might this be another symptom of the hypersensitivity?
The yellow portion of the graph in the Audio Setup Wizard is what is called the noise floor. It is a measure of inherent (ambient) electronic noise and background noise if you're not speaking. The green bars are the strength of your voice as compared to the ambient noise. All microphones exhibit some level of yellow. The very best tend to have a very low noise floor while dictating. Therefore, when we see a low yellow noise floor while performing the last step in the audio setup wizard, we can almost be assured of a good microphone being used.
Oh yes, the battery box is the standard one you need to make the NC-95 work. As far as I can tell, it's not making the signal any louder. Or at least it's not supposed to be doing that.
Electret condenser microphones used in the typical computer headset require a small amount of voltage to vibrate at the sound of your voice. Many microphones get the voltage from either a USB soundcard or a soundcard/sound chip. However, on the Andrea microphones when used with Sound Blaster sound cards require a battery box to provide the bias voltage.
Martin Markoe, eMicrophones, Inc.
See us at: http://www.emicrophones.com
I would say the green spike
I would say the green spike when you shake the cord indicates a bad connection rather than putative "hypersensitivity". I do get slight, minimal yellow signal fluctuation with my Sennheiser ME3 from ambient noise. But even if I flail the devil out of the cord I can scarcely see any fluctuation in the signal indicator.
You need to get a completely new mike to try out. That means not re-using the box either.
I wonder where you can find a microphone vendor? Hmmm . . . Let me ponder that a bit.
Bruce
ara_vartanian wrote:Okay, I
Okay, I have something. First of all, I noticed that my yellow bar is almost always fluctuating up and down, even if there is absolute silence in the room (even when I'm holding my breath -- my breath seems to make no difference). It is usually around half way to the distance at which point it would be a green bar. For those of you who have the optimal arrangement, when there is silence all around, is the yellow bar nowhere to be seen? Second, shaking the wire at any point causes the bar to land firmly into the green. However, the closer to the microphone itself that I jangle the cord (as opposed to the closer towards the computer), the greater the disturbance Is this supposed to be happening? When I play that back, I hear a giant crashing sound quite disproportionate to the meager sound of my stirring the cord. Might this be a source of the problem? Or might this be another symptom of the hypersensitivity?
Thanks again for all your help.
Oh yes, the battery box is the standard one you need to make the NC-95 work. As far as I can tell, it's not making the signal any louder. Or at least it's not supposed to be doing that.
Without going into great technical detail, based on what you're saying, it seems obvious that you have a bum microphone. Shaking the microphone cord should have absolutely no effect on the noise floor (i.e., yellow bars). However I will add this for your edification. Each one of the vertical bars that you see in the graph represent a particular frequency range. Low frequencies to the left higher frequencies to the right. The general range from left to right is from 0-100 Hz to 5000 Hz. Therefore, it is sometimes helpful to pinpoint where the yellow bars peak within this range. In the lower or left portion, the problem is coming from the lower frequency ranges and so on up the line.
Chuck Runquist
Former DNS SDK & Senior Technical Solutions PM for DNS with Lernout & Hauspie (L&H)
Chuck, I'm not sure what you
Chuck,
I'm not sure what you mean by "yellow bars"? When Ara mentioned his "yellow bar", I think he was referring to what I see on my Dragon bar just to the right of the turn on the microphone icon? On my Dragon bar (at the top of the screen) my yellow bar goes from top to bottom but varies in width as I speak until it turns green about a third of the way across. I don't have any display that looks like a histogram. What were you referring to?
Jim wrote: I'm not sure what
I'm not sure what you mean by "yellow bars"? When Ara mentioned his "yellow bar", I think he was referring to what I see on my Dragon bar just to the right of the turn on the microphone icon? On my Dragon bar (at the top of the screen) my yellow bar goes from top to bottom but varies in width as I speak until it turns green about a third of the way across. I don't have any display that looks like a histogram. What were you referring to?
We are talking about the last step in the Audio Setup Wizard. The yellow (bottom of the graph) is the noise floor. This is noise inherent in the system as well as background noise.
Martin
Okay, I have more perplexing
Okay, I have more perplexing data for everyone. It turns out that I think I can solve my problem of the hypersensitive microphone, but at the expense of something in Dragon not quite working very well. You see, as a previous poster just said, it seems as if this microphone the Andrea NC-95 requires the APS-100 battery box to work with the SoundBlaster Live! soundcard. For kicks, I tried unplugging the battery box. It turns out that my dictation perceives just fine and the ambient sound being registered by my yellow bars is only a fraction of its former self. So why did I ever need the battery box to begin with? Well, when I check my acoustic settings, I can't seem to pass the speech quality tests without the box. It consistently fails. Because of that, at first, I supposed that I needed the battery box. This doesn't seem to be the case. I can get acceptable dictation without a "leaky" microphone, but now my speech quality tests consistently fails.
Even without the battery box, shaking the cord does cause the yellow bar to go into the green. So perhaps I still have a bad microphone, regardless of these other issues. But I think it is odd that the speech quality tests consistently fails even though my dictation is fine.
Thanks again for everyone's help.
I would just like to add as
I would just like to add as an addendum that the problem did seem to be my microphone. I've switched to a Sennheiser microphone and both the strange dribbling and the phantom commands have disappeared. They did seem to be related. As words dribbled, Dragon seemed to gravitate towards options on the menu bar. I don't fully understand Dragon's behavior here, but the Sennheiser's rock steady.
Its great that you've solved
Its great that you've solved your problem! The Sennheiser headset is great, ergonomic gear.
To clarify, I presume to speak for others and say that we think the old mike was generating noise from electrical interference which DNS attempted to interpret as speech data, either dictation or commands.
Bruce