Hi Peter,
The sound source was just some white noise from a small speaker, general ro=
om tone, and urban roar in the background. I picked sections without obviou=
s anomalies like closely passing cars or planes. I think the consistency of=
sound source is not all that important, since the relative differences bet=
ween capsules still should remain the same. I ran the test several times an=
d that seemed to hold true.
If my set of ten was at all representative of the typical spread for the bu=
lk em172, 70% are fairly tightly grouped, 20% are somewhat outside one dire=
ction the other, and 10% are another good step outside. I did not compare c=
ombinations by ear except between the one highest and one lowest of the gro=
up, and in that case the difference between overall sensitivity and compara=
tively skewed frequency response curve was highly noticeable by ear. It is=
important to know that luck of the draw could make or break a balanced pai=
r.
Thinking over my method, the reference capsule could just as well be one fr=
om the testing set - it would just become the arbitrary zero line in the di=
fferences graph. Only the relative values (the differences between the diff=
erences) are important.
Mike's method would of course be a lot simpler. Checking resistance values =
was actually the first thing I tried, but at the time I did not notice the =
correlation with my differences tests. I may have stopped searching once I =
found my answers.
John Hartog
rockscallop.org
--- In Peter Shute <> wrote:
>
> John, what kind of sound source did you use for the recordings?
>
> How important is this matching process? I bought four EM172s and used the=
m all in my SASS without considering trying to match them.
>
> Peter Shute
>
>
> --------------------------
> Sent using BlackBerry
>
> ________________________________
> From:
> To:
> Sent: Fri Apr 05 06:11:02 2013
> Subject: [Nature Recordists] Re: Matching electrets
>
>
>
> Hi Anders,
> I am not sure what the best or easiest way is but this is what I did to m=
ake pairs from ten em172 capsules - the capsules came with leads already so=
ldered on.
>
> Number each capsule with a marker to keep them straight.
> Make a test fixture - stereo cable with screw terminals to easily swap ou=
t capsules.
> Record a stereo recording of each EM172 capsule beside a single reference=
capsule. In this case for the reference capsule I used an EM158 I had lyin=
g around.
> Open each recording in your audio editing software (I used Adobe Audition=
1.5), run a frequency analysis of a section of each, copy the data to the =
clipboard, and paste it into a spreadsheet (I used MS Excel.) You may need =
to adjust the frequency analysis resolution to limit frequency divisions to=
a manageable number of data points. I do not know if "copy to clipboard" i=
s a common feature with other editing or analysis software.
> In the spreadsheet, subtract the left (reference) channel from the right =
(EM172) of each set to find the differences in frequency response.
> Plot the points to make a graph showing the differences from reference of=
all ten capsules, and use that to decide the best pairing.
>
> It will be interesting to hear how others have done it.
>
> John Hartog
> oregonsoundscapes.com
>
> --- In <naturerecordists%40yahoogr=
oups.com>, "kronseth" <anderslangstrand@> wrote:
> >
> > Hi everyone!
> > Sorry if/that this question has been raised probably infinite times.. S=
o I have received a batch of electrets, both omnis and cardioids.
> > The ones I have bought before have come in matched pairs, but this time=
I=C3=82=C2=B4ll have to do it myself.
> > I understand there a quite a few ways of doing it. Anyone has a preferr=
ed method?
> >
> > Best
> > -Anders
> >
>
>
>
>
>
>
|