Topic: Are match pairs really that critical

[dactylus]

^
Thank you DSatz!

[datbrad]

I can say from personal experience that if one microphone in a pair has a different sensitivity level than the other, odds are good that the response curves are going to look different as well. It's true this will not always be clearly audible in a recording of the output from a PA system at a distance. However, it can be noticeable at the time of making the recording. Behaviors of the signal that make the recorder levels seem uneven, causing adjustment, can quickly dance from left to right. This is caused when one mic is more sensitive to certain frequencies than the other. What looks like great stereo separation can actually be surges from the variations between the mics taking a particular note that should be represented in both channels equally to be louder in one mic, which distorts the imaging. This is why matched pairs are not necessarily sequential in serial numbers. When "matching" mics into stereo pairs, mfgs are seeking two units that have the same open circuit voltage, and response curves that match as closely as possible. Users can by chance obtain two single mics of the same make/model that match closely and as long as the sensitivity matches, that's usually going to be fine for PA taping, imo.

[John Willett]

I can say from personal experience that if one microphone in a pair has a different sensitivity level than the other, odds are good that the response curves are going to look different as well. It's true this will not always be clearly audible in a recording of the output from a PA system at a distance. However, it can be noticeable at the time of making the recording. Behaviors of the signal that make the recorder levels seem uneven, causing adjustment, can quickly dance from left to right. This is caused when one mic is more sensitive to certain frequencies than the other. What looks like great stereo separation can actually be surges from the variations between the mics taking a particular note that should be represented in both channels equally to be louder in one mic, which distorts the imaging. This is why matched pairs are not necessarily sequential in serial numbers. When "matching" mics into stereo pairs, mfgs are seeking two units that have the same open circuit voltage, and response curves that match as closely as possible. Users can by chance obtain two single mics of the same make/model that match closely and as long as the sensitivity matches, that's usually going to be fine for PA taping, imo.

With some manufacturers, microphones will have serial numbers that are quite a way from each other - this is because the mics are matched after production.

Neumann matched pairs normally have sequential numbers because the serial numbers are put on the microphones *after* they are matched.

Sennheiser MKH 20/30/40 series are manufactured to a very high tolerance that they don't need matching.  However, the MKH 8000 series have glued capsules to get the same 16mm diameter diaphragm into a 19mm diameter housing - these are more difficult to manufacture and do need to be matched.

Microtech Gefell do not charge extra for matching, providing the two microphones are ordered at the same time and pair-matching is requested at the time of order.

But all these are high quality manufacturers and you are likely to find that random samples of Schoeps mics are better matched than a matched pair of inexpensive Far East microphones.

[DSatz]

The threshold of perception for differences between microphones, when you're using just a pair of them for stereo recording, is much, much lower than it is in the most common studio application--close or moderately close multiple miking of multiple sound sources, where the output of many microphones will ultimately be mixed into a single stereo channel of the final result.

Whenever any microphone manufacturer can guarantee that all its products are within a total of about 1/4 dB in the midrange, and about 1/2 dB at the top and bottom, then I will agree that there's no further point in pair-matching that company's microphones. Since I don't expect to see that happen during my lifetime, I will gladly pay a little extra for pair-matching for any microphones that I'm going to use as coincident or closely-spaced stereo pairs. (Or if it's offered for free, I'll take that, too.)

--best regards

[voltronic]

Whenever any microphone manufacturer can guarantee that all its products are within a total of about 1/4 dB in the midrange, and about 1/2 dB at the top and bottom, then I will agree that there's no further point in pair-matching that company's microphones.

This may be slightly off-topic, but there should be an international standard of some sort for what constitutes a "matched pair".  Even among the top-quality brands, I'm sure they all have different internal standards for this, and a matched set from one manufacturer may be a completely unacceptable variance to another.  This probably goes hand in hand with the overall tightness of manufacturing tolerances and QC.

I think the numbers suggested above would certainly be a good goal to shoot for, but some of the big names might have trouble meeting that level of consistency (while others will meet that level without breaking a sweat).