Skip to main content

Hi there,

I have got a doubt about phase in a XY stereo configuration. I've always heard that it's a very mono compatible configuration. But a technician here in my city always says that we should avoid this configuration, because if the two micros are not exactly matched (And by that he means a factory preset XY configuration only), there are severe phase problems in the high frequencies.

Is it true? Sometimes I need to use this configuration better than a semicoincident one, but I have this doubt. Could someone help? Thanks!


TheJackAttack Mon, 11/22/2010 - 10:23

All you need to know about XY is to get the mic capsules as close together without touching as you can and aim them correctly. Also, the 90 degree angle is a starting point. I have modified this based upon the spread of the ensemble without much issue. If I have to change too much though then I choose a different pattern to use. There are no more phase issues with this pattern than there are with M-S or ORTF.

As to matched, that can be a can of worms. Even a "matched pair" from a factory will become "unmatched" over time as the microphones age. As long as they are closely similar and of similar age I wouldn't get too uptight about that. Quality control being what it is today, as long a I purchased a pair of microphones at the same time that were very close, and I used them equal hours in equal environments, I have not seen much issue at all. As a reference, I own matched pairs of microphones and pairs that I have matched myself for that purpose. I think it is useful if the microphones are serialized to keep notes about which mic was in what position and what pattern was used so that if you do notice anomalies you can adjust accordingly.

Big K Mon, 11/22/2010 - 10:35

Hi, ...
I use XY for ages with identical mics, but all pairs I used are not what is called matched. If the positioning is correct there should not be any phase problems. At least, I have never heard anything like that. Even with ORTF config, which places the mics appart, I get good results.
Microphones of the same type and brand have very small tolerances and with XY config their membranes are only a few millimeters apart. So, the acoustic events hit the membranes very much at the same time, which gives you this reasonable mono compatibillity. If you place them too close to a wall, aiming towards, e.g., a drum set, you might get reflections that produce problems, but that would be not a matter of matched or not matched mics.
Afaik, matching a pair of mics has more to do with finding 2 mics with extremely similar frequency response and output. Such a matched pair sounds indeed better in a critical situation. For the majority of recordings in XY AND when using good mics, in the first place, it is not really show stopper to use an unmatched pair.

anonymous Mon, 11/22/2010 - 11:47

thank you for your answers. By "Factory preset configuration" I did'n mean a matched pair, but a stereo mic with its capsules in XY or two mics sold as mounted pair in XY. That is what this technician always says, that it's very hard to set up manually this stereo configuration without having milimetrical errors that cause phase issues.

I've used this configuration sometimes and I didn't heard phase issues between the mics. What he says sounds really weird, but this technician is an experienced one, he works for the state radio and he records usually a lot of live concerts and other music, and maybe he can hear phase issues in the very high frequencies that I cannot.

Thank you!

TheJackAttack Mon, 11/22/2010 - 11:55

The difference of a few mm is not going to cause phase errors. Neumann has many stereo microphones that have a movable top capsule that are still valued in the thousands of dollars as does AKG. There are many fixed patterned microphones as well in the high end without any issues-ribbon, tube, condenser.

The low end microphones that are fixed capsule should still not exhibit phasing issues because of position differences of a mm or two. They have other issues in regards to their electronics but that is a separate area. Do you hear phasing when you use your particular mic? Your bigger problem with fixed capsule positions in an all in one microphone is positioning in relation to the ensemble. Too close and you won't have a big enough sound field. Too far and you might have a good depth of field but be too far from the ensemble. Too close to a wall is the same regardless of fixed or not fixed.

Just my thoughts. Big K might have different ones than me.

apstrong Mon, 11/22/2010 - 18:13

BobRogers, post: 357296 wrote: Well the wavelength of a 10,000 Hz sound wave is 3.4 centimeters, so "milimetrical errors" don't cause phase issues at any frequencies not geared toward the all important cocker spaniel market.

According to a frequency chart I'm looking at, the wavelength of E9 is 33mm, and F9 is 31mm. It's only a semitone, but... so would a difference of a few millimetres change the pitch slightly in the highest ranges (like 10-15K especially), with the effect becoming larger as the frequency increases? If so, does this affect things (noticeably)? Or is that just part of the magic of harmonics? Personally, I've never heard any problems mic'ing in x/y configuration, except that I can never get it to sound wide enough for my tastes and often use a spaced pair instead. Woof woof.

TheJackAttack Mon, 11/22/2010 - 18:33

Our ears don't hear all the harmonics equally no matter what and as if that weren't bad enough, they aren't even close to one another! If the position of the capsules or the "matched" aspect of the capsules mattered that much then every stereo pair would be astronomically expensive and fixed into position at the factory. Like Big K, I've used plenty of pairs of mic's as coincident or near coincident pairs without them being matched with no unwanted phasing or comb filtering issues. For the recording work my company does (classical location), I do now try to have very similar microphones for coincident pairs.

As to frequency length mattering, when we hear a note "in tune" with another it is a function of how their harmonics mesh. The 2nd partial of e9, which most people can't even hear, is half the wavelength or twice the frequency. That's fine if we are talking a 2:1 octave but usually not. For a P4 interval for instance we are now talking a 4:3 ratio which means for e9 and a9 to sound in tune the harmonic tone in question is e11. Now we are in cocker spaniel range. Worse, there are many folks to whom the top note of a piano is a pretty nebulous thing and that is only c8.

As to freqs above 10k, a recent hornlist argument resulted in someone creating a sine wave melody of the freres jacques them from Mahler 1 at 12K and someone else posted one at 15K. The top note turned out to be about D9 on the former and B9 on the later. Many folks couldn't even hear the notes above 11k and less than a quarter could hear all the notes of the higher. We sound engineers often get obsessed with freqs up to 20k when most of these are felt rather than heard. I'm not saying they don't matter altogether, but in current society the numbers of people that can even feel pressure of 15khz freqs is diminishing rapidly.

It's Belgium Triple night so forgive meandering, buzzed typing. And if you've never had the Trappist delight........well more for me.

Boswell Tue, 11/23/2010 - 03:10

The key to what our Unregistered Guest's friend is saying may be that he works for a radio studio. Encoding stereo for FM pilot-tone transmission tends to emphasise phase problems in the original L-R source, and that is why many radio engineers say that M-S microphone techniques give better results for stereo radio.

Keeping the M and S channels separate (instead of L and R) right up to the pilot-tone encoders at the transmitter is practicable really only for single-transmitter local FM stations, and only then if they make an effort to make it happen. It doesn't actually cost any more, other than a few L-R decoders where monitoring in L-R is required, but I doubt the stations use it considering the commercial and other pressures they are under.

BobRogers Tue, 11/23/2010 - 04:22

apstrong, post: 357308 wrote: According to a frequency chart I'm looking at, the wavelength of E9 is 33mm, and F9 is 31mm. It's only a semitone, but... so would a difference of a few millimetres change the pitch slightly in the highest ranges (like 10-15K especially), with the effect becoming larger as the frequency increases? If so, does this affect things (noticeably)? Or is that just part of the magic of harmonics? Personally, I've never heard any problems mic'ing in x/y configuration, except that I can never get it to sound wide enough for my tastes and often use a spaced pair instead. Woof woof.

Different mic positions don't cause pitch shifts. They cause time shifts - delays. A one millimeter change in mic position causes about a .003 millisecond delay. So that does cause phase problems, but at VERY high frequencies. But as Boswell points out, the engineer was concerned encoding these signals as FM, and the carrier frequencies are in the MHz range. So there could well be problems that don't crop up in regular audio recording.

Big K Tue, 11/23/2010 - 04:39

XY not wide enough... If you adjust the central axis of each mic to the edge of the sound sources and do this in a distance to it that allows the pick-up sensitivity/ polar pattern
of your mic to still cover the middle as strong as the center, you get the wide picture possible without using more mics. If you close the middle axis in the picture gets more
narrow, if you go beyond the left and right edges it gets more narrow, again, as well. XY give you sharper localization of sources, A/B gives you more room depth at less localisation.
ORTF combines both aspects to a certain degree and under certain circs.

My dearest, most versatile stereo mic I have worked with ( and still is not in my arsenal...yet..):
http://www.thomann… SCHOEPS MSTC 64 UG - U.K. International Cyberstore
Drawback: capsules are not adjustable, so, not always optimal within small rooms..

What I mean with :" too close to the wall": if you position the XY mics with the backs to a wall ( XLRs almost touching ) and the wall is reflecting the signal back to the mics ( usually not coming directly
to the mic's back axis, which would be greatly attenuated by the polar pattern of the mics ) it can cause a sort of comb-filter effect.

Radio guys often have other technical priorities and tend to make up problems where there are none..
What I can vouch for: they have the same poor hearing all humans BUT they have meters and scales telling them there should be something wrong....
Of course, you need to be bat to hear it or you can sense phase problems of FM and AM carriers with a raised wet finger..

As ever: use your ears. If you don't hear it and it sounds is very likely to be ok... If in doubt, switch to mono and listen again ( if needed ).

Hello Jack!!
We should author our answers and remarks We could safe time...
You take upper part and I scribble the lower...

Big K Tue, 11/23/2010 - 04:51

TheJackAttack, post: 357309 wrote:
It's Belgium Triple night so forgive meandering, buzzed typing. And if you've never had the Trappist delight........well more for me.

Man, I love the Rochefort No. 10 with its 11 %. I should not be saying that as a Bavarian, but I am too much of a beer connoiseur as that I could ignore the pleasure of enjoying some of the Belgium strong brews. I had a mild (over-) dose of it regularly ( not for the aclohol, though !!) when I worked in the Netherlands, quite close to some of the best beer brewing monestaries...
Have on for me, too, mate.


Have u deleted your post?

Boswell Tue, 11/23/2010 - 10:04

BobRogers, post: 357343 wrote: ...But as Boswell points out, the engineer was concerned encoding these signals as FM, and the carrier frequencies are in the MHz range. So there could well be problems that don't crop up in regular audio recording.

I know we are getting quite technical here, and that this is the Open Mic forum rather than the DIY Pro Audio forum, but it's worth pointing out that the carrier frequency in question is that of the difference channel at 38KHz. This uses a suppressed carrier technique (to minimize the energy above the normal audio band), and is regenerated in the correct phase by the receiver's stereo decoder from the 10% amplitude 19KHz "pilot-tone". The L-R (difference) information is double-sideband amplitude-modulated about the 38KHz carrier, and so the "audio" signal that is put into an FM transmitter has a bandwidth of about 53KHz. The 19KHz pilot-tone slots into the small gap between the top of the L+R (mono) baseband and the maximum extent of the lower sideband of the modulated difference (L-R) signal.

The phase concerns at the audio level are brought about by this sum-and-difference subcarrier method of encoding the stereo information, and would be similar if the wideband audio were put straight into a local stereo decoder and not sent out as an FM radio signal. Of course, there are other corrupting effects in radio such as noise imparted by the FM transmission and reception process. Supplying the stereo encoder at the FM transmitter with audio directly from Mid (sum) and Side (difference) channels would help reduce the effects phase discrepancy and other related problems when modulating the difference signal. The stereo decoder in the receiver automatically performs a process identical with M-S to L-R decoding.

RemyRAD Tue, 11/23/2010 - 11:33

When you think of phasing issues you are really talking about filtering. Equalizers in analog consoles are really nothing more than phase manipulators. It's a type of phase shift that they create that we call equalization.

Purchasing actual stereo microphones is more of a convenient setup than using all sorts of hardware to mount your microphones. I frequently utilize MS microphone technique which, depending upon my microphone selection i.e. U87's, 414's, Beyer M 160/130, SM 81's. All require different mounting hardware which also requires a certain amount of creative capabilities.

I got a good LOL out of your Radio guy's information. I came across a similar situation while making a recording in New Zealand. It was an operatic recording for Los Angeles record label. We were in a huge cathedral in the heart of downtown Wellington. We had our sessions in the morning & afternoon. After us & later in the afternoon & evening an operatic recording was also being done by the New Zealand radio guys for a New Zealand opera singer. Our biggest problem was, we wanted to leave things set up. And they believe that one could only get a good recording in this cathedral with one of those AMS multi-capsule stereo microphone. My producers didn't want that and neither did John M. Eargle. I was going to use a MS microphone for front center. John was the record labels primary engineer but because of his schedule, he couldn't do this job. He gave me his pair of SANKEN CU 41's and asked me to space these cardioid microphones on either side of the conductor. And I used numerous other highlight microphones. The New Zealand radio guys said they had been recording in this cathedral for 40 years and that my microphone setup scheme would yield bad results. However, the New Zealand record producers really liked the way my microphone setup was and didn't exactly like that single AMS Sound Field microphone. As a result, the New Zealand producers for the New Zealand record requested my setup over the New Zealand radio guys. And they gave me credit as recording engineer on this other operatic recording that I really had nothing to do with. Needless to say, the New Zealand radio guys weren't very happy about this because " they know better". Obviously not. All of this is so subjective and while I mic'd up most everything the way I wanted, I was slightly skeptical about John M. Eargle's primary left-right microphone setup request. And although these microphones sounded good, I really felt that they were a gimmick microphone? Why did I feel that way? Because there is not a single capsule in this fixed cardioid microphone. There are 2 capsules like a speaker has. It had a small capsule & a larger capsule both coupled together like a woofer & tweeter. I actually confronted the SANKEN microphone people at the following AES about their gimmick microphone. They took great offense to that and told me it was not a gimmick. But then I asked him about crossover notch distortion between the two capsules? To which they replied that there wasn't any. That's because it probably wasn't exactly utilized like woofers & tweeters with the SDC crossing over to the LDC but perhaps were both wideband & summed together without any crossover network/electronics and its associated distortion artifacts.

So really anything goes if it sounds good to you. What I was doing didn't sound good to the New Zealand radio guys but did sound good to my Los Angeles-based producers and to the New Zealand producers. So the 40+ year professional New Zealand radio guys were upstaged by the 30 something-year-old hotshot American engineer. And no one died. We actually came up with a reasonable compromise. I had come equipped with my dual API 3124m's. The New Zealand radio guys utilized a AMEK BC1 portable mixer. My API's work cleaner and more aggressive sounding in comparison to the AMEK. But hey, COMPROMISE is not a dirty word. I just didn't think that my recording and the sheen that I liked from my then brand-new API 3124m's. In the end, it really didn't make a difference and I still garnered a Grammy award nomination for my recording.

So while you are futzing around trying to make sense of XY microphone technique I just want you to know that radio guys are radio guys and not recording guys even though they do a lot of recording. There are fundamental differences between getting IT on the air or getting a Grammy. The New Zealand radio guys recording would have been fine as well with their AMS SoundField microphone if you happen to like that kind of sound. And although it was much more phase coherent than my microphone technique was it was my bigger wider spread of sound that both the Los Angeles producers & New Zealand producers liked. The AMS microphone was too "MONO-MATIC" tasting to me in spite of its phase coherence. And John M. Eargle's spaced cardioid request did not give me the solid center that I particularly like from MS microphone technique.

So you see, your decision-making process is moot. And inexpensive single head XY stereo microphones is a convenient way to go for someone like yourself. Audio Technica & numerous others make such stereo microphones which are affordable. But when you really think about it, as a recording engineer, I personally would rather have 2 discrete microphones instead of a single stereo microphone. This gives you so much more versatility. So while your radio guy friend is not completely wrong he's also not completely right. Radio guys over many years all seemed to get a little crusty. While most of us recording guys sit on the edge of technology. So if you need a fast setup and you needed to get it on the air fast, I would go for your radio guys suggestion. Otherwise, half the fun with being a recording engineer is working it to get the sound you want with whatever it takes. Convenience be damned!

Land of the living damned
Mx. Remy Ann David