how to match electret mic capsules ?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

isophase

Well-known member
Joined
Jul 2, 2007
Messages
240
Location
Lutece
hello,
can't seem to find any answers by doing a search...
title says it all.
i'm building a pair of omni mics with panasonx Wm61A capsules and i was wondering how do i match them ?
capacimetre? ohm-meter ?
thank you
 
hi Keith, thank you for the fast response !
i had never heard of the Pistonphone before so i've learned something today. respect.
assume the capsule doesn't have an "internal preamp" or FET, would you still need a Pistonphone to mesure something out of it ?
i guess there is mechanical constraints involved like the tension of the actual diaphragm and maybe also the magnetic remanence of the electret material ? if you cut the trace that connects to ground on one of these small electret capsules would you then be able to mesure something valuable like a fixed capacity ?
 
Use a very well-matched stereo preamp, headphone amp, and headphones.

Put one capsule into BOTH channels. Set up a selection of acoustic sources: boom-box, best friend reciting a poem from memory, fan. Trim the stereo so the sound is *exactly* in the middle of your head. (Few headphones will do this perfectly.)

Put one capsule on each side. Capsules side-by-side and touching. Repeat.

If the sound is still *exactly* in the middle of your head, you are done.

If the source is now off-side, trim your balance control so "most" of the sound is on-center. Listen to the extreme highs and lows. If they are off-center, you can't easily match this pair. Replace one capsule and try another.

When/if you do find a fairly well-matched pair, note the balance adjustment needed. Personally, I'd fix it in post-processing. I knew my mic A was 2dB hotter than my mic B, and I ran a 2dB gain-adjust the next day when preparing the track.
 
In more ambiant recording  [ classical ] or room mics
fixing later will increase also any room or system noise as well
once in a trk, 2 db may be nothing , but still better
to go to the source or fix it there as much as possible
it could be maddening thinking it's your positioning when
it's really the microphones

good question , can you match cheap capsules ?
 
On the recording end with stereo 2 track, if it's possible I'll try to center a test tone speaker output and record a bit of it for level reference.  Sometimes helps make decisions about steering after the fact, and confirming or denying preamp gain or mic differences in the result.  Sometimes it's totally useless. 
 
thanks for the replies,
i'm thinking of putting two capsules standing as close as possible in front of a speaker and pass some white noise throught them and monitor with a spectrum analyzer + listenning tests. i guess this should do ?  
for now i tested my prototypes using an HDR7506 headphone set and a cheap mixer. One of the capsules was about 3dB less sensitive than the other, i then swapped the capsules around with their respective preamps and still got 3dB less on that same capsule... So its the capsule.
i will have to buy more of them as i only bought 4 and one was destroyed by doing some mounting tests for my microphone body prototype, two have the linkwitz mod (the ones with the 3dB difference) and one was kept original so i could make comparison tests against the linkwitz modified ones. so now i'm stock and have to buy more capsules to make an acceptable pair. though i am quite tempted to make the linkwitz mod to my last original WM61A and see if i can better match the gain with one of the already modded capsules...
also, i have a pair of (exellent) Avenson Audio ST02 i did some comparison tests against, i was not surprised to find out that it sounded quite different from the mics i've built (with and without the Linkwitz mod), the ST02 seems to lack bottom end when compared to my prototypes (or my microphones are generous with low frequencies either way) also the ST02 have a tiny bit less gain but by ear the noise/hiss seems to be about the same.
i have no idea what capsules the ST02 use but i'm quite familiar with these mics as i've used them a lot (and i like them a lot) and by experience i almost always use a high pass filter when i use them (close or room mic) so i'm very curious to test my mics in real recording conditions and see what's going on with the very generous low end it seems ?  
BTW, the circuit i'm using is the "standard" phantom powered design with the 2N5087s transistors found on the net.
i will post some photos of my proto soon :)
i would like to thank Stefan from http://www.homemademicrophones.com/ for maximum inspiration, Tape op who also inspired me with their omni mic project , and of course the great Lab members that take the time to answer my (apparently good) question ;-)
thanks
J
 
 
> fixing later will increase also any room or system noise as well

If mike noise is below room noise (as we prefer), then the fix corrects the unbalance.

If mike noise is audible above room noise (often true for $1 electrets in quiet rooms) then of course the result is unbalanced mike hiss. Then what you really want is two QUIET capsules with tolerable response differences.

> center a test tone speaker output and record a bit of it for level reference.

Good theory. Sometimes good sanity check. May even reveal some major goof before the show starts.

IME, the act is never perfectly balanced across my mike (or my mike was not centered on where/how they were going to play?) and plain live 2-track always sits off-center on tape (or WAV). Sometimes that was the conductor's intent. Sometimes the viola feels a bit tired, the violin sits too far forward, my level pot has some-dB mis-track, and it all leans the same way.

I dunno what tools others use. CoolEdit 2000 will run an average over a sample. Here's 0:02.00-0:08.00 of XEX's Svetlana (1980):

------------ Left Right
Peak Amplitude: 0 dB -.6 dB
Possibly Clipped: 3 0
Maximum RMS Power: -3.77 dB -6.35 dB
Average RMS Power: -8.88 dB -10.36 dB
Total RMS Power: -8.57 dB -10.05 dB
Using RMS Window of 50 ms

The left channel is 1dB hotter on-average than the right; and the peak unbalance is comparable. Over the whole 2:19.7 track the average difference is 2dB (there is a 0dB peak in each channel, but more possibly-clipped sample in left than right). Because the track is pioneering Techno pasted-up from scraps of 1/4" tape onto 8-track, there's no reason to expect "balance", and I'm mildly stunned it is as close as it is.

On choir, I would re-balance to <2dB over passages with no piano (with accompanying piano off to the right we must unbalance the final stereo track to keep the star choir "centered").

Carley's Simon's You're So Vain, with strong center voice and professional mastering, appears to have 1dB unbalance. ABBA's Dancing Queen appears be centered within 0.3dB in parts and 0.02dB overall (I may be working from an over-processed MP3).
 
Sure , and basses [ string or voice ]
are never centered
I get my best balences when the conductor
[ for choirs ] moves people around

It's simular to changing levels during the recording
where too much of a difference will be noticable in " room tone "
there's where the rear channel may be handy
for being there
 
you guys must be talking about Large AB stereo mic set up, in that case the stereo image is so large its difficult to reproduce anything clear in the center, in that case ok, a perfect match is not that important but how about with a jecklin disk ? hahahah nerdy
well in either case i think i would prefer to have a pair of microphones that at least sound calibrated to my ears !

edit: yes i ment a Jecklin Disck not dick, oups. :)
 
> how about with a jecklin dick ? hahahah nerdy

Did you mean: jecklin disk?

And why would that matter? Only the wide-space pair is "immune" to small unbalance (because it never really comes into focus and is riddled with interchannel phase oddness).

Gain differences are the EASIEST correction, and gain-setting is just basic audio craft.

With low-cost capsules trimmed to same "overall" level, you must then check for irregulaties. The bass extension is set by air leaks, and full-bass capsules have ultra-small vents. One may be drooping at 30Hz and another with some glue in the hole may droop at 80Hz. More to the point: the diaphragm is "tuned" like a cymbal to resonate at the top of the audio band. Zildjian craftsmen hand-hammer each cymbal for the desired resonance. $2 capsules put the diaphragms on slap-dash. They aim for the average resonance to be, say +1dB at 13KHz, but the plastic film has thick and thin areas, so some come out 11KHz and others 15KHz with several dB variations. Some show several resonances which may be from different causes and have different errors.

_I_ think you wire two capsules to good headphones (verified with mono) and listen for the whole spectrum low--high to stay dead-center. I think the ear is a better judge than a spectrum analyzer.

> basses [ string or voice ] are never centered

Well, at the end of Don Juan the statue of the dead commodore starts moving and singing deep, telling Juan that he's going to hell, then yanks Don into a burst of fire. That's such a great scene that it is usually staged DEAD-center.

But I know what you are saying.

> I get my best balences when the conductor [ for choirs ] moves people around

Much of my salary was live before a paying audience. The act was balanced for the crowd. If possible.... you can't have the soprano solo run-down the risers from her spot in choir, and in the narrow Chapel stage she could not run-behind and get out front. In such situations you can't balance the act for the mikes. Just do the best you can to balance for the crowd, maybe shift your mikes, and maybe re-balance later.

If you have true diverse stereo field (not panpot stereo), quite large unbalance sounds very natural. Performers often don't comment on it even when I know there is 5dB average difference.
 

Latest posts

Back
Top