Live Recording & Matched Microphones

[artto]

What is it about matched pairs of microphones that makes the

recording sound better?

Background: over the past 20 years I have produced over

forty live-in-concert recordings for a local Chorale. Only two of these were

made with a matched pair of main (L+R) mics. The first one was recorded a few

years back with a pair of rather expensive Neumann which I borrowed. But “that

sound” has never left my head. Recently I bought a pair of matched omni

capsules for my Oktava mics. And there was “that sound”.

All of this got me thinking. Many (if not most) commercial

recordings of orchestral and chorale music tend to use multi-mic techniques.

And most of these recordings (IMHO) lend a somewhat strident sound to higher

frequency instruments such as violin and piccolo. Recordings that are least

offensive to me are often recorded with 2 or 3 matched main mics such as Mecury

Living Presence or Telarc. I’m not absolutely sure, but I doubt that the

multi-mic’d recordings used (for instance) 5, 6 or more “matched” microphones

to obtain that screechy midrange and treble sound quality.

This last spring concert sounded absolutely terrific with

the matched pair. The violins retained that “woody” quality with no stridency.

The low was full with lots of weight and bloom. The pipe organ notes go down to

the center of the earth kind-of-thing. No shrillness in the alto sopranos.

I wonder if the something similar would happen if all the

microphones in a multi-mic setup were matched.

[Arkytype]

Artto,

While having "matched" recording microphones is a laudable goal, I think there is a more important factor to consider when using multiple microphones and that is the time arrival differences.

If you are using a pair of main L & R microphones plus some extra mics for orchestral "enhancement", a mono sum during mixdown may reveal some nasty aural "combing" or stridancy as you have noticed.

With the advent of high quality capturing and editiing software, it would be easy enough to record each microphones on a separate track and then "time slip" the accent microphone tracks to match the main mics time-wise. Of course, you have to decide where in the soundstage you want the listener to be seated!

As I have access to high-quality tube/solid state Neuman mikes, a new hig sample rate digital multi-track recorder and Pro Tools software, I am making it a point this fall to record as many University of Arkansas bands/orchestras as possible to experiment with "time coherent" recording. We typically use eight microphones in recoprding but the final mixed product may only incorporate the L & R plus two or three other tracks for sweetening.

Back in the mid '70s, Paul Klipch & I recorded the Arkansas Sympathy Orchestra (not a misspelling--they were pretty bad) when they played in Little Rock. I used a mid-side (M-S) recording technique with a Neuman U-47 and U-87 while Paul preferred two widely spaced omni AKG mikes.

The primary advantage of the M-S setup is that the arrival time to the microphone diaphrams is identical. The disadvantage is that the placement of the two microphone array is very critical in order to get a good balance. Wide spaced omnis always seemed to have a lack of bass to my ears.

Lee

[MikeFord]

Matched is just an idea without numbers. Most professional quality microphones have a lot of consistency within brands and models, so I'm not sure what would be gained by trying to find two that have slightly closer response curves than two picked at random. IMHO what matters is where the microphones are located. Stunning realism is possible using fairly cheap microphones if placed properly. Stunningly bad is possible even using top quality microphones in less than ideal locations and orientations.

[artto]

Lee, you lucky dog you! You got to do some recording with PWK? [] [Y]

I must admit I also prefer the spaced L+R. I've recorded this Chorale (sometimes with a small orchestra or other guests such as a large group of hand bells) with different techniques over the years. I haven't used M-S yet but I would sure like to try it if I ever get a figure-8 mic or the adaptor for my Oktavas.

We've been recording in the same room now for many years (College Church, Wheaton, IL). Wonderful acoustics for this sort of thing. I have to mix and record live-in-concert, one take, no going back. This also creates some logistical problems with mic placement. I know where the best place for the mains are given the conditions I have to deal with. The Musical Director has eventually understood this and how much it affects the recording quality so that now I usually have some say as to where soloists or musicians are located.

I'd like to know of your experiences with "time coherent" recording when you get around to it. I believe I have an older OEM version of Pro Tools which I doubt has the ability to "time slip" any additional mics. I've been using Sound Forge Pro for quite a while. I'm using Digital Audio Labs Card Deluxe recording at 96Khz/24bit and the sound is very good. Only one card at the moment so I'm limited to 2 track stereo. I'm not sure Sound Forge can do the time slip thing. I guess I'll find out when I get a second card to handle four tracks.

Interesting that you find the spaced omnis lacking in bass. I've experienced just the opposite.

My thinking is that possibly the matched frequency response and sensitivity alleviates some of the abberations caused by time/phase interferences (combing). For instance, if one mic is + 2dB at x Hz and the other mic happens to be -2dB at the same Hz there should be more comb filter effect than if both mics were within 1 or 2 dB of each other. All I can say is that the matched pair sound way better than the unmatched ones (same brand and model). And its the same thing I experienced with the match Neumann a few years back.

[artto]

Matched is just an idea without numbers.

Most professional quality microphones have a lot of consistency within brands and models, so I'm not sure what would be gained by trying to find two that have slightly closer response curves than two picked at random.

IMHO what matters is where the microphones are located. Stunning realism is possible using fairly cheap microphones if placed properly. Stunningly bad is possible even using top quality microphones in less than ideal locations and orientations.

I'm not sure what you mean by "just an idea without numbers". All the matched pairs I've seen come supplied with graphs of the response curves with actual numbers of their performance.

I couldn't agree more that mic placement is of paramount importance. I've made some incredibly "live sounding" recordings with some rather inexpensive Audio Technica high impedance mics. But when recording live-in-concert I don't always have much choice as to what kind of mic setup I can use or perfect placement for everything. As time moves on and the Director hears the differences in recordings, I've been given more say as to stage setup.We even left this facility (College Church) for a while but the new auditorium had dreadful acoustics, especially for recording, and I managed to convince them to get back to College Church. And besides, the church has a really nice pipe organ too.

[MikeFord]

My thinking is that possibly the matched frequency response and sensitivity alleviates some of the abberations caused by time/phase interferences (combing). For instance, if one mic is + 2dB at x Hz and the other mic happens to be -2dB at the same Hz there should be more comb filter effect than if both mics were within 1 or 2 dB of each other. All I can say is that the matched pair sound way better than the unmatched ones (same brand and model). And its the same thing I experienced with the match Neumann a few years back.

I think rather than matched pair, in a professional setting you would call it meeting specifications, there just should not be that much variation in response compared to location and orientation.

I don't think frequency response would have a material effect on time or phase issues, and any effect they might have could just as well be positive with a mismatch and negative with matched.

In some brands, it used to be common for all of production to be checked and then sorted into two or three grades which were sold to different channels. A matched pair could be from any grade, and to meet any standard, which is why my comment was so negative. Two matched consumer grade microphones would not be my first choice, vs anything from a professional line.

[Arkytype]

Artto---

I don't know if any of PWK's recordings from the mid-seventies have survived; that's a question for Jim Hunter. Mine (half-track 7.5 ips Dolby are hopefully still in strorage. Unfortunately, the Tandberg deck is long gone. BTW, PWK used a half-track high-speed (7.5 -15 ips) Revox B-77 with no noise reduction and the aforementioned AKG phantom powered omnidirectional capsules..

While matching frequency response between mikes is a given when we speak of "matched", the idea of phase matching is usually not considered. Bruel & Kjaer used to sell matched laboratory-standard measurement microphones. They were matched within a half dB or so FR-wise and within a few degrees phase-wise. The matched pair was placed in a jig that had them pointing nose-to-nose with a calibrated spacer in between. This setup was used used in sound intensity measurements which necessitated the close FR & phase matching.

While their mikes are, (to quote PWK), "As flat as a book keeper's arse.", their signal-to-noise ratio is inversely proportional to their FR. A one-inch diameter capsule has the highest S/N ratio and is flat out to about 8 kHz. Their 1/4" diameter capsule is essentially flat to almost 100k Hz but the S/N is 20 dB or so worse than the one-inch. Danish Pro Audio (DPA) mikes was a marriage of B & K technology and higher S/N ratio products with XLR connectors instead of the 6-pin proprietary instrumentations mikes. They are highly sought after and command high after-market prices on E-Bay.

While I haven't measured the "comb filtering" effect when combining spaced mics, I suspect it is as pernicious as having multiple identical drivers in a loudspeaker cabinet. Our university department has a large conference space and setting up multiple mikes with an omni-directional sound source shouldn't be too hard to do to replicate recording an orchestra.

To play the Devil's advocate---one could say, "When I am sitting in the audience and there are 80 or so instruments playing, the direct sound from each arrives at my ears with 80 different arrival times!"

Even if all 80 members played in unison say a 1/16th note (tempo of 60 beats per minute), I think our marvelous sense of hearing would process the 80 direct paths and the subsequent multiple reflections and not hear multiple "combing" of the individual 250 msec sources.

Looks like I'll be busy this fall proving or disproving the ill effects of combining multiple microphones!

Lee

[Quiet_Hollow]

The most realistic recording of an orchestra would use a matched pair of mics in a binaural in-the-ear arrangement (only to be played back on headphones) or a matched pair mic setup that cut the line of sight to the "virtual listener" prescribed in the Dolby arrangement for 2 channel listening. It's purely a matter of superposition. To blend more than 2 mics for a 2 channel recording will introduce a whole lot of instrumental phase problems that the mixing engineer will have to sort out. ie. Why try to mic a xylophone when the conductor has beaten the percussionist over the head enough times to make them sound just right. In fact, if you wanted to make a heck of a recording, try placing the microphones at the podium...to hear what they hear.

[MikeFord]

Its really hard to mess up with a two microphone setup. Radio station I did some engineering at did all the smaller live stuff with a simple pair of microphones, not matched, just same model and brand, in a 120 degree fixture that mounted to a stand.

I forget the name, but there was a popular 3 mic stereo recording technique as well.

[Arkytype]

One advantage of the Mid-Side recording technique is the ability to vary the stereo stage width from mono to full left and right separation. Since the capsules are coincident time-wise, there is no mono summing frequency-dependent phase canceling.

One can record the two mic signals on separate tracks and adjust the stereo image width in playback or just mix the two while recording which is what I do using a Tektronix 760A Stereo Audio Monitor to set the "fur ball" pattern.

I don't think frequency response would have a material effect on time or phase issues, and any effect they might have could just as well be positive with a mismatch and negative with matched.

Keep in mind that any change in a loudspeaker or microphone's transfer funciton will have a commensurate change in phase.

Lee