[Scott Dorsey]

Jerry Steiger wrote:
"Scott Dorsey" wrote in message
...
B&K makes an artificial ear system and an artificial mouth that are
specifically designed for testing these things. They try very hard to
simulate off-axis responses of actual human body parts. They aren't
cheap, but they are surprisingly effective.

Sorry, my answer only confused things even more! Our products aren't used
like normal voice cell phones. It would be like Maxwell Smart using his shoe
phone. The smallest are about one pound, the largest over two pounds. So I
guess the first would be like Maxwell using 99's shoe and the second would
be like him using the gentle giant's shoe. We actually planned on putting a
speaker in a smaller one so it could be used like a normal cell phone but
decided it didn't make sense for most of our customers. If we ever did that,
we might need to look into the artificial ear and mouth. Certainly we would
need to do considerably more testing to meet the certification standards of
the cell phone companies.

The artificial ear and mouth are just perfect for that sort of application;
they get used for walkie-talkies all the time, as well as headsets. But they
_are_ intended for close-talking applications.

The microphone is used to record field notes or possibly for voice commands.
The speaker is used mainly for notifications and the normal computer beeps
and squeaks. I suppose people could play music over it to (thats how we used
to test them), but there are much smaller systems that would sound much
better and most everybody already has one, just like cell phones.

The thing about the artificial mouth is that it will do a good job of
modelling someone talking into it, both in the correct manner and from
various different directions. If you need real intelligibility measurements
to show someone, this is critical. If you don't need objective measures
of intelligibility, and you can assume that folks will always be positioned
in the right place, you can just use a headphone driver for testing.

Sometimes the unit is mounted on a tripod with surveyor's instrument. In
that case it is roughly three feet off the ground and three feet away from
the surveyor's head.Sometimes it is mounted a pole, putting it a little
farther from the ground and closer to the surveryor, say 4 feet up and 2
feet away. Sometimes it is just held in a persons hand, so it could be as
close as a foot and a half or so. They are always used with the display and
keyboard pointing fairly close to straight at the user, so off axis response
beyond 30 degrees or so isn't very important. I figured I would just measure
the response straight on and not worry about off axis response, although I
could easily rotate the units under test if it became important.

If people are using the thing from three feet away, you need to be able to
measure response with a sound source three feet away. Because it may be
very different than the response up close, especially if you are using a
directional microphone.

My guess is that under typical surveying environment, at three feet away
from the mike the environmental noise problem is going to be severe. And
anything you do to make the mike more directional is going to make wind
noise problems worse.

Incidentally, let me commend the Gentex omni capsules to you. They are
very resistant to rain and mud. They also sell noise-cancelling mikes
for close-talking use, but that does not sound useful to you.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."