Matched Pairs don't matter

Sometimes maybe it doesn't matter too much, with real good mics made under tight quality control.

My pair of Shure SM81's for example can't be bought as a matched pair, but Shure claims they're all more or less matched, and that's fine with me so I bought a couple and love them.

But to say that it NEVER matters is wrong. Look around these boards for comments on the Octava MC-012's.
 
My pair of 603's are a bit different. one feels slightly "scooped" compared to the other. Is a very minimal difference, but something that I picked up on in the first few uses of the mics. So far it hasen't been an issue.
 
I found a noticeable difference in my 2 mxl 603's. They were purchased at the same time, though not as a "matched pair". So far, it hasn't been an issue for me.


Twist
 
I have 4 Marshall MXL603S's (and a stereo pair on order) and 1 of the 4 603S's I have has a lower gain and sounds a little different than the other 3. I'd like to hear what Harvey has to say about this, I understand he has a box full of MXL603S's, so he should know.
 
Mics are quirky pieces of equipment. No manufacturer can guarantee absolute consistency for every mic, so they have a "window of acceptance", whereby any mic that can pass thru that window is considered "good" and anything that doesn't fit inside the window is rejected.

Imagine a 3 foot by 3 foot window, wide open, and you can see that a 35" circle would pass thru the window, but so would a 35" by 35" square. Both go thru the window, but they are certainly different from each other.

Humidity, machine adjustments, thickness variations, component tolerances, and even the mood of the assembler that day all play a part in determining the consistency and final sound of any one single mic. The differences we hear between "identical models" of a microphone can be broken down as follows:

1. Self noise
2. Frequency response
3. Sensitivity

In actual order of importance for matching, it would look like this:

1. Frequency response
2. Self noise
3. Sensitivity

Usually self noise can be separated into 3 broad categories:

1. Higher than normal
2. Normal
3. Lower than normal

After a manufacturer has broken the mics up into these three categories, they try to match the frequency response. That breaks down into about 5 categories:

1. Neutral
2. Slight bass boost
3. Slight treble boost
4. Slight mid boost
5. Slight treble AND bass boost.

What about dips? Well, a slight bass boost is close to the sound of a slight treble cut, so they sound pretty similar. A cut mid sounds similar to a slight treble and bass boost. Anyway, that gets the mics split up into 15 general categories.

They then look for mics in each pile with similar sensitivities, and they wind up with a bunch of matched pairs.

Now of course, the "matched pair that YOU would like to get would have low noise, neutral frequency response, and high sensitivity, but that's not the definition of a "matched pair" - it's a definition of the best "matched pair" in that particular batch.

Okay, on to your question, DJL. I'd send them all back to MXL and ask if they can find a mic that matches better to your 3 similar units. I'd also make sure the 3 similar units were clearly marked.
 
Thanks for the response, all. Harvey, I've got a question, based on your order of importance. If the sensitivity is the least important of the 3 characteristics, then is the Studio Projects "colored dot" method of "matching" mics worth much, to those interested in matched pairs? The way I read your response, if you have two mics with matched sensitivity (which is what the colored dots are for), then they're really only quasi-matched at best.

BTW, I posted the quote because I wanted to write to Sweetwater and let them know that the Tech Tip really sounded like bunk to those in the know. Now, I feel like I can say that with confidence. Thanks!

-mg
 
Sensitivity can be easily matched using the channel trim pots, if the noise levels and frequency response are similar.

Frequency response is the most critical area, since any major differences will result in a screwed up image.

Let's assume two mics are identical in every way, except one of the mics has a narrow 3dB peak at 500Hz and the other has a dip at the same frequency. You have the mics set up in stereo and the guitar player plays a chromatic scale on the guitar.

Whenever the guitar player plays the note that corresponds to 500Hz, the image will jump over to whichever speaker is connected to the mic with a peak in it. It will shift 6dB to one side or the other. That can be very disconcerting.

Smaller differences are far less noticable and mic frequency matching within a dB or so is pretty damn good.

How companies match mics will vary from company to company. If their noise levels and frequency response are pretty consistent, sensitivity matching is all that's needed. I believe Earthworks considers frequency response to be most critical, and they will custom match pairs to within a 1/2 dB.

Personally, I'm most concerned with similar frequency responses, followed closely by similar self noise levels (but the similar self noise consideration would be less critical for drum overheads than for quiet acoustic guitar work).

As with most things, there's no one easy answer to this stuff.
 
Harvey, I want to thank you for jumping in and helping me and all those who read this thread. I have a much better understanding about "matched pairs" and the best matched pair of the batch. And, as soon as my back-ordered stereo pair of MXL603S's get here, I'll send the other four 603's (marked) in and ask for some better matching. Did you get your MXL603S's matched at Marshall Electronic's, and or what? Thank you very much for your time and help. :)
 
harvey is right on, as usual. how noticable these differences are depends on the material being recorded, the ears of the listener, the overall mix/style of the music, etc.

i would tend to have some doubts about the "matchedness" of a pair of mics from a heaper manufacturer in particular. aside from that, let your ears do the walking.
 
Harvey Gerst said:
Humidity, machine adjustments, thickness variations, component tolerances, and even the mood of the assembler that day all play a part in determining the consistency and final sound of any one single mic.
Click to expand...

Hmmm. Anyone know of any hookers we can have sent on over to the assembly lines? I'll get a hold of the prozac and ecstasy. :D
 
DJL said:
Harvey, I want to thank you for jumping in and helping me and all those who read this thread. I have a much better understanding about "matched pairs" and the best matched pair of the batch. And, as soon as my back-ordered stereo pair of MXL603S's get here, I'll send the other four 603's (marked) in and ask for some better matching. Did you get your MXL603S's matched at Marshall Electronic's, and or what? Thank you very much for your time and help. :)
Click to expand...
Yes, mine were matched by Marshall, for free. But in all fairness, when I went to the last AES show in Los Angeles, I explained a very simple, low-tech method (which I came up with years ago at JBL) that they could use to achieve accurate matches very quickly. They were very skeptical when I described it, but they at least tried it, and they found it speed up accurate matching dramatically, like 1/10 the normal time.
 
Harvey Gerst said:
Yes, mine were matched by Marshall, for free. But in all fairness, when I went to the last AES show in Los Angeles, I explained a very simple, low-tech method (which I came up with years ago at JBL) that they could use to achieve accurate matches very quickly. They were very skeptical when I described it, but they at least tried it, and they found it speed up accurate matching dramatically, like 1/10 the normal time.
Click to expand...

So, please tell us. :)
 
Actually, it's a technique I developed at JBL about 35 or 40 years ago that I'd rather not explain in detail just yet. I will give you some background though. At the time (mid 60's), I was head of Quality Control at JBL.

We had to provide the Army with matched pairs of 075 drivers (the ring radiators - I don't know the current name for them). It would tie up the lab for the entire day, running curves and trying to find the identical units out of a batch of 50 or so units.

While testing them initially to make sure they worked and weren't rubbing or buzzing, I noticed there was a signature sound to each driver when listening to white noise. Just for fun, I separated them into batches, by their sonic signature, and wheeled the cart into the lab. I told the lab technician that these units sounded similar and I pointed out the pairs that I thought sounded identical.

To everybody's amazement, the pairs I thought were identical really were - they matched to within about 1/4 dB. So were all the other pairs that I had matched by ear, but the "similar" units matched to within 1 dB.

The next time we got an order from the Army, I did my "matching trick" again, and after running the charts, we wound up with the same results. I could match a batch of 50 drivers by ear in less than 30 minutes as well as the chart recorder in the lab, which took all day to run and match the drivers, using the paper chart recorder curves.

The Army didn't ask for our curves; we shipped them as "matched units", and they ran their own confirming tests. So, from then on, we supplied the Army with "Harvey matched" drivers, and simply didn't tell them how we matched them. Of course, to make sure, I'd do one "by ear" match from the tray of drivers and have the lab confirm they really were matched.

Over the length of the Army contract, it probably saved us hundreds of man hours and it freed up the lab for other projects.

We'd still leave the tray of drivers in the lab all day, since Ed May (the chief engineer), Bart Locanthi (head of R&D), Gus (the lab technician), and I decided it would probably be best if it remained our little secret.
 
Harvey Gerst said:
Actually, it's a technique I developed at JBL about 35 or 40 years ago that I'd rather not explain in detail just yet. I will give you some background though. At the time (mid 60's), I was head of Quality Control at JBL.

<snip>
Click to expand...

Of course you know my mind is wondering why you'd rather not give the details for a simple quick mic matching test. But, I now have some ideas to try with some white noise. Thank you again.
 
DJL said:
Of course you know my mind is wondering why you'd rather not give the details for a simple quick mic matching test. But, I now have some ideas to try with some white noise. Thank you again.
Click to expand...
Mainly, it's a bargaining chip that I can use with some companies to either enhance my credibility, or to possibly sell as a technique. Some of my ideas are still commercially marketable, I think.

Some of my ideas for different products are just now appearing in products coming on the market, and I'm currently in negotiations with some manufacturers on some new products I'm designing. While I try to share as much general knowledge as possible with all of the people here, I hope you will understand that I prefer to keep any of my propriatary ideas under wraps for my own personal use.

I believe I've already posted a simple technique here in this forum for matching mics at stores.
 
it seems to me (and i am an amateur, so don't kill me for suggesting it) that it should be possible to check matching by close-miking a simple source out of phase and monitoring the output with the right equipment.
 
mgraffeo said:
Harvey, as always, thanks.

for those looking for the "matching mics at stores" thread, here it is:

http://www.homerecording.com/bbs/showthread.php?threadid=38175&highlight=matching+mics+at+stores

-mg
Click to expand...
Thanks for thak link. I remembered writing something about testing mics in stores, but I forgot what it was that I wrote. Going back and re-reading it, damn, that's not bad.

At 66, I'm getting Alzheimers and Deja Vu at the same time; I think I've forgotten this before. :(
 
You must log in or register to reply here.
Back
Top