Topic: how do they match mics?

[Depechemode1993]

Just reading the site and they were talking about matching mics. How do you go about doing this? Can a consumer do it themselves or does the mic company or a retailer fix this for you? why would the specs not be the same for every mic of the same model? I was just curious because if I ever get mics again I am probably going to save up for some MK4's or get some real nice DPA's for open taping.

[Walstib62]

The 1 st stage output of the mic preamp gain is set to equal output based on a controlled input to the capsule.

[stevetoney]

The 1 st stage output of the mic preamp gain is set to equal output based on a controlled input to the capsule.

Many mic circuits do not have voltage gain nor provide any feature for adjustment.  The most critical aspect of matching is frequency response, not sensitivity (although a matched pair should match for both).  Matching is generally done by selection, although with tighter production tolerances on the microphone capsule that becomes an easier task.

An end-user could do it given a sufficiently large sample of microphones, the sample size depends on the capsule tolerance for that particular model.  Ten mics ought to be sufficient for most brands.  It can be done in a store with relatively little gear if you have access to a mixer and a patient salesperson.  It's faster and easier to do if you can record a repeatable full-range source.

So, as a followup to this answer, if you're thinking about buying a pair of mics that don't have plots, but the mics have consecutive (or close) serial numbers, does that really tell you anything other than there's a potentially greater chance that they're closer than say two mics with serial numbers that are far apart?

[NOLAfishwater]

well first, each microphone capsule has to set up an online dating profile and then........

[stevetoney]

I don't know if that's true, but I do know that the SM81 I resurrected for a friend after her kid dunked it in the toilet wasn't going to match any other SM81 . . .

Funny you mention.  I've been debating people in another thread who seem to think it's OK to leave a mic unprotected from the rain when you're recording an outdoor show.

[Church-Audio]

The 1 st stage output of the mic preamp gain is set to equal output based on a controlled input to the capsule.

Many mic circuits do not have voltage gain nor provide any feature for adjustment.  The most critical aspect of matching is frequency response, not sensitivity (although a matched pair should match for both).  Matching is generally done by selection, although with tighter production tolerances on the microphone capsule that becomes an easier task.

An end-user could do it given a sufficiently large sample of microphones, the sample size depends on the capsule tolerance for that particular model.  Ten mics ought to be sufficient for most brands.  It can be done in a store with relatively little gear if you have access to a mixer and a patient salesperson.  It's faster and easier to do if you can record a repeatable full-range source.

So, as a followup to this answer, if you're thinking about buying a pair of mics that don't have plots, but the mics have consecutive (or close) serial numbers, does that really tell you anything other than there's a potentially greater chance that they're closer than say two mics with serial numbers that are far apart?

This depends on the manufacture of the mics in question Neumann is going to have much better quality control then say Shure. The main reason why you guys love one mic over another has less ( in some cases ) to do with the over all quality then the predictability of the mics in question. Predictability comes from two possible sources..


A- Quality control and testing of each capsule with a philosophy that capsules that don't measure up get destroyed.
B- Better build quality insuring less "sub standard" capsules are produced and thus not as much need to match capsules.
C- Luck sometimes companies get lucky and end up being able to produce a great product with very little testing.
D- Company does not know what quality control is and regularly sells a shit product.

The best type of company to go with is in my opinion Company A. This company will charge more for the product but in the end you will end up with a more consistent product that gives you a predictable sound every time you use one. So when we are talking about consecutive serial numbers we must first talk about who's serial numbers they are. And where the company falls under A, B, C and D. The other thing to take into consideration especially when we are talking about mics.. Where is the serial number located? Is it on the mic body? If so that has little to nothing to do with the mic capsule it self as you can have to serial numbers on two bodies mics that are from two different batches of mic capsules. BUT if the serial number is on the capsule we can assume that the capsules came from the same batch of "stock" and should sound very similar. The only exception would be when a manufacturer does the serial number stamp at the end of the manufacturing process. Then you have absolutely no guarantee of anything. So generally speaking serial number matching with anything but very high end mics with serial numbers on the capsules is not any guarantee of any sort as to how close they will sound to one another.

Chris

[su6oxone]

A bit off topic but does Schoeps always provide plots for factory matched mics or do they include the certificate sometimes?  Wondering because my matched CCMs came only with a certificate.

[illconditioned]

Here is a quick-and-dirty way to test capsules, especially small and/or omnis.

Play pink noise through your speaker at high volume (eg., 90dBA).  Wear earmuffs during this, because it will sound crazy loud, like standing behind a waterfall.

Put both capsules side by side about 15-20cm in front of the speaker and record a stereo track.  Then play back with real time spectral analysis.  I use the "Waves PAZ Analyzer" plugin in Wavelab.  Set it to stereo and slow response.  The curves should line up.  They may deviate in the high end, say above 15kHz or so.  They may also deviate at the low end, below 50Hz.  But they should line up really well at points in between.

This is a good sanity check too.  For example, before selling a pair of AT853 capsules I'll do this to make sure one is not damaged (or worn out somehow).

Not scientific, but it works...
  Richard

[wklitz]

all I know is I bought one Schoeps MK21 back in Dec '07 and another just last week, obviously not a "matched pair", but they certainly are close enough to pretend during my testing.

[Kindguy]

well first, each microphone capsule has to set up an online dating profile and then........

My Neumann's did this. They are very happy now.

[Depechemode1993]

soo... lets say I buy some MK4's or even some DPA 4022's, If I buy two mics what are the chances that they are pretty matched? What happens if they are not? It seems like it would be an easy process in post to make them even, but would you be able to return a mic if it happens?

[su6oxone]

soo... lets say I buy some MK4's or even some DPA 4022's, If I buy two mics what are the chances that they are pretty matched? What happens if they are not? It seems like it would be an easy process in post to make them even, but would you be able to return a mic if it happens?

Schoeps mics are said to be pretty close to each other even if not officially matched, especially I would think when the serials are close.  My non-matched Schoeps MK4 caps are close in serial numbers and within a dB or two from my experiences in post processing.  Returning a mic/cap would depend on the return policy of the shop you are purchasing from, so B&H might be a good idea for MK4 since they are a non-special item and they won't charge a restocking fee. 

[Depechemode1993]

B&H is probably the best store ever.

[DSatz]

When Neumann, Schoeps or Microtech Gefell talk about matched pairs, they really mean microphones with selected, matching pairs of capsules. They don't adjust microphones to match each other; they pick and choose pairs from among a production batch. Schoeps will select matching capsules (or CCMs) of any type on request, for a fee; for those models which Neumann sells as matched pairs, they typically choose amplifiers that have consecutive serial numbers as well--but that's mainly a vanity thing. When not specifically selected for a match, consecutive serial numbers don't even indicate that two items are from the same batch, let alone that their characteristics match.

With professional condenser microphones that have sonically neutral electronics (basically, any modular series with different interchangeable capsules), all the amplifiers of one given type should have essentially identical frequency response, while the gain should be within ca. 1 dB for the better brands whether they're from the same batch or not.

The greater variable is the capsule. Being electromechanical it has "a foot in both worlds" so to speak. Capsules vary both in frequency response (particularly at both ends of the range) and in sensitivity, and the extent of this variation generally exceeds the variation among amplifiers.

Sometimes two capsules match well in frequency response but not as well in sensitivity. In such cases it helps if you always put the more sensitive capsule onto the amplifier with the slightly lower gain, and vice versa.

The characteristics of capsules change somewhat over time, so a pair of capsules that was considered to be adequately matched (say) five years ago might not meet the manufacturer's matching criteria now. Unfortunately that's a fact of life.

--best regards

[illconditioned]

Previous posts in this thread have painted the picture quite well. Pair matching really means selection--microphones aren't normally trimmed to match one another. With professional condenser microphones that have sonically neutral electronics (which includes any series with different interchangeable capsules), all the amplifiers of one given type should have essentially identical frequency response, and the gain should be within 1 dB for the better brands, whether they're from the same batch or not.

The less controlled variable is the capsule. Being a device that converts sound energy into a varying electrical voltage, it's not purely electrical; it's electromechanical, and has "a foot in both worlds" so to speak. Capsules vary both in frequency response (particularly at the extremes of the frequency range) and in their 1 kHz sensitivity.

When Neumann or Schoeps talk about matched pairs, they really mean microphones with selected, matching pairs of capsules. For those models which Neumann sells as matched pairs (which is not necessarily the same thing as a stereo set--some of their stereo sets aren't matched), they typically choose amplifiers with consecutive serial numbers as well, but that's mainly a vanity thing, and everybody should realize it if they don't.

Sometimes two capsules match well in frequency response but not as well in sensitivity. In such cases it helps if for the more sensitive capsule, you always use the amplifier with the slightly lower gain, and for the less sensitive capsule, use the amplifier with the slightly higher gain.

--best regards

Any comments on my rough matching idea... just looking at the (slow response) spectrum in real time?  I've done a few lavalier mics this way and got good results.  If nothing else it is just a sanity check, to make sure two differently-sourced AT853 capsules are working.  I'm not matching down to +/- 1dB or anything, but I'm just looking for gross differences in performance.  As I've said earlier, one should check all electronics gear and verify that it is really working!

Oh yeah, I've noticed large variations if different mic types.  For example Countryman B3 and Nevaton MCE400 had quite large variations, not only in sensitivity but in the curves themselves, while other mics, like Sanken Cos-11 and Countryman B6 and DPA4060 were extremely close, the curves line up almost exactly.  A common mis-match is to have mics matched at 1k or so, but the responses diverge smoothly with increasing or decreasing frequency, one going up gradually, the other going down.  I think the manufacturing process has something to do with it it, and maybe electrets too.  Could be sensitivity to temperature, soldering, packaging, ?  I don't know.  The guys who build these things probably know, but we're entering the realm of secrecy here I think...

  Richard

[guysonic]

One technique I find most useful in matching mics (this is done here for matching DSM mics) is to apply pink noise to two closely positioned mics at >1 meter distance from a speaker. 

Amplify the output of the two mics and invert the signal to one of the mic channels, then sum the two channels to mono and observe the spectrum graph of this mono signal.

If the two mics are perfectly identical over the pink noise frequency, then a straight line results when viewed on a spectrum graph display.  Where the sum of two mic signals (one channel inverted) are not an identical match, the summed output appears as a deviation from a straight line over the test frequencies.

If one mic has higher output than its mate, then a straight line still occurs if mics have perfect "bandwidth over frequency match," but the straight line will have some value other than 0 db indicating amount of overall output sensitivity mis-match.  This sensitivity mis-match is quite OK when known and can easily be corrected with simple level adjustment of recorded channels in post edit process.

While omni mics need not be rotated for doing a test like this, directional type mics should be rotated 30, 90, 180 degrees to observe any change in off-axis response match existing.

[Giant_Rick]

That's a very interesting thread, goes straight to my Favorites on Google Chrome!

I have two questions.

1. Is there an equivalent of the 'PAZ Analyzer' for Audacity? Or may I just use EAC (after making a mono track of the stereo one -- what does 'invert' mean, why do I have to do this and is there a way to do this with Audacity?)?

2. Can soldering have any effects on the capsules?
I've ordered some Panasonic WM61-A to make a small set, but I'm afraid that the soldering operation can be harmful. Obiviously I'll touche the mike the least time I can with the soldering tip.

Thanks!

[bucsab12]

2. Can soldering have any effects on the capsules?
I've ordered some Panasonic WM61-A to make a small set, but I'm afraid that the soldering operation can be harmful. Obiviously I'll touche the mike the least time I can with the soldering tip.

This may sound like a stupid comment but make sure that you use some sort of heat sink for the capsule when you are soldering the wire to it.

I have used these small "crocodile" connectors (I am not sure if that is the correct term in English, please excuse me) when I soldered the capsule to the cable and it worked great. The capsule stayed at the same temperature, maybe a little hotter.

One time, when I forgot to use the crocodile connector, the capsule became very hot after touching it for only about 2 seconds with the soldering iron...

[Giant_Rick]

This may sound like a stupid comment but make sure that you use some sort of heat sink for the capsule when you are soldering the wire to it.

I have used these small "crocodile" connectors (I am not sure if that is the correct term in English, please excuse me) when I soldered the capsule to the cable and it worked great. The capsule stayed at the same temperature, maybe a little hotter.

One time, when I forgot to use the crocodile connector, the capsule became very hot after touching it for only about 2 seconds with the soldering iron...

I have some iron alligator clips that helped me last time I made a work with two cheap and crappy capsules.
They didn't have any solder on.

Do the WM61 come from some solder on, and do I have not to use any other solder but the one there's already on?

[Giant_Rick]

They do have solder on their pads, but I always freshen it with a tiny dab of mine.

I'll do the same.
Thanks for replying!