Omnis in XY. Does it make sense?

[davidvaldes]

I´ve been told by an engineer that he uses a pair of Earthworks QTC 40 in a XY configuration.

As long as I know, you need time difference or a pressure gradient (or both) to get stereo. Because those mics are omni, there´s no pressure gradient (maybe on high frecuencies, where they get more directional but, what happens on the low end?) and, because XY is a coincident technique, there´s no time difference.

He claims he gets fantastic results, but I can´t see the physics behind a XY configuration with omnis.

Am I missing something? Could you please elaborate on this subject?

[Marlan]

It would essentially sound like an out of phase mono recording.

[Audiop]

Originally Posted by Marlan

It would essentially sound like an out of phase mono recording.

No, why?

It would basically sound like a mono-recording with some slight stereoinformation in the top octave.

X/Y with (particulalry small) omnis makes no sense.


/Peter

[Corran]

Earthworks actually suggests in their manual that you should try X-Y recording with them.

I am stumped as to why but I've been meaning to try it, just because. Obviously with another pair running just in case...

Does your engineer friend have a demo made this way??

[davidvaldes]

He claims that technique is described in the booklet accompaning the microphones...

Honestly, I´m really disconcerted, as I can´t see how omnis in XY can produce stereo.

[davidvaldes]

Does your engineer friend have a demo made this way??

Will try and see if I can post a sample.

[rumleymusic]

That is very strange, I can't see how it would provide a very good image. We are talking about QTC40's here, right? Earthworks does make a few cardioid mics.

[Suda Badri]

2 omnis in XY will definately give you spatial cues... your middle will be well re-enforced, the sides nicely diffuse... it sounds great... might not make sense but it sounds the part... plus, you can take a pair like that closer to the source since no proximity and general omni yummyness, very natural stereo... just dont expect the usual width and you will be fine...

worth a try...

[davidvaldes]

TO make things even more complicate... He uses them in a vertical axis, in a down/up configuration instead of the regular left/right one. He uses that technique for classical guitar.

[Corran]

Well now a single instrument definitely is different than a large ensemble. I wanted to try it on a wind ensemble or orchestra.

Daniel, yes Earthworks mentions the X-Y technique specifically in their omni mic manuals.

[Suda Badri]

Originally Posted by davidvaldes

TO make things even more complicate... He uses them in a vertical axis, in a down/up configuration instead of the regular left/right one. He uses that technique for classical guitar.

Wait, What? so he pans the top facing mic to the left and the bottom to the right or vice-versa? or does he not pan at all... in which case it wont be a stereo recording... unless you tilted your head to listen to stuff or it would sound like the guitar was played upright...

[ysmgj]

Originally Posted by Suda Badri

2 omnis in XY will definately give you spatial cues... your middle will be well re-enforced, the sides nicely diffuse... it sounds great... might not make sense but it sounds the part... plus, you can take a pair like that closer to the source since no proximity and general omni yummyness, very natural stereo... just dont expect the usual width and you will be fine...

worth a try...

+1.

or try ortf, it will give you the width you're looking for.

again like suda said, might not make sense in ortf either but it doesnt sound bad and you can get rather wide,... but it'll essentially be an angled spaced pair in this case.

[rumleymusic]

Well ORTF omnis (really just closely spaced AB) is understandable. I use that quite frequently for solo flute.

I am beginning to understand here.
For XY, saying it is for a guitar or other single instrument makes much more sense. It is essentially a mono-'ish' recording with a primary center image and a sense of ambiance due to off axis variations that you cannot get from a single mic. The up-down spacing makes sense on guitar if you are trying to very slightly get a L/R image based on frequency.

Wouldn't try it with a large ensemble though.

[ysmgj]

^^^ cheers for explaining the up/down bit.

[Suda Badri]

Yeah thanks, Man I gotta try this whole up/down... sounds cool!

[Celloman]

I read that Bruce Swedien calls this "fat mono". An interesting sweetener mic technique.

Mike

[Washington]

Originally Posted by davidvaldes

Honestly, I´m really disconcerted, as I can´t see how omnis in XY can produce stereo.

Well, it wouldn't be a very large and obvious stereo to start with, and that's exactly what could be interesting. I love M/S to dial some subtle width to acoustic guitars, while retaining a strong center presence and most of all avoiding the dreaded "10-ft-wide instrument". Could be interesting on drums too, giving a little dimension to cymbals on an almost mono kit (love that sometimes...). Got to give it a try, thanks for the idea.

[JonesH]

Originally Posted by Celloman

I read that Bruce Swedien calls this "fat mono". An interesting sweetener mic technique.

Mike

Darn it, I thought I did something original when I started doing this
I used a X/Y config with two 4006-TLs with 50mm APEs in close proximity to a pop acoustic guitar to get a slightly more 3d image. It worked fine for the song, but I haven't tried it in bigger acoustics.

[spicemix]

It's a noise reduction technique.

If the area of a circle is pi r squared, halving r is going to reduce the area quite a bit, non? So a 1/4" capsule is going to pick up far less signal than a 1/2". Given the mic's preamp self-noise is relatively constant, those tiny cap mics are noisy.

Putting two together so that the signal will be in phase doubles the signal level by doubling the capsule area. Still not as big as a 1/2" capsule, but the noise is uncorrelated and thus will partially cancel itself out. Leading to greater signal to noise ratio.

They tell you to do it but don't tell you why, because they don't want to admit the laws of physics mean their mic is noisy. Those mics do have advantages of course: SPL handling, frequency and transient response, etc. Use them as close mics.

[Suda Badri]

Peter... You da man...

[plus6vu]

I've been using Earthworks mics for about 10 years and love them.

I tried the XY technique with omnis once on a 40 piece wind ensemble. Never tried it again. It was essentially mono. Those mics are very omni, so putting them in close proximity to each other, unless close miking, won't get you much of an image.

Now, I would use it on a single instrument if I had an occasion to do so. I do that all of time with other mics, but normally cardioids and panned close together.

ymmv
s

[Marshall Oliver]

Originally Posted by Suda Badri

2 omnis in XY will definately give you spatial cues... your middle will be well re-enforced, the sides nicely diffuse... it sounds great... might not make sense but it sounds the part... plus, you can take a pair like that closer to the source since no proximity and general omni yummyness, very natural stereo... just dont expect the usual width and you will be fine...

worth a try...

So with a with this technique, would it be right to say that would have the focus of a cardiod pattern, without the proximity effect?

Thanks in advance....


-1-

[Mark A. Jay]

Originally Posted by spicemix

It's a noise reduction technique.

If the area of a circle is pi r squared, halving r is going to reduce the area quite a bit, non? So a 1/4" capsule is going to pick up far less signal than a 1/2". Given the mic's preamp self-noise is relatively constant, those tiny cap mics are noisy.

Putting two together so that the signal will be in phase doubles the signal level by doubling the capsule area. Still not as big as a 1/2" capsule, but the noise is uncorrelated and thus will partially cancel itself out. Leading to greater signal to noise ratio.

They tell you to do it but don't tell you why, because they don't want to admit the laws of physics mean their mic is noisy. Those mics do have advantages of course: SPL handling, frequency and transient response, etc. Use them as close mics.

If we assume that the inherent electrical random noise is present in both mics and in equal power, then yes, as they are uncorrelated sources, the overall random noise signal power (due to the mics' electronics) increases by 3 dB. If the random noise components are not of equal power then the result will not be a 3 dB increase, but can be calculated by first measuring the noise floor of each, and doing a log-sum.

As far as the acoustics go, in an X-Y configuration the mics experience nearly identical incident pressure as a function of time (due to near-identical location), but this is particularly true for low(er) frequencies. As such, and assuming the sensitivity of both mics is basically the same, then there will be a 6 dB increase when they are summed, but mind you, not for all frequencies - because at a certain point frequency-wise they will no longer be in lock-step phase-wise (because the mics occupy two different places in space they cannot see the same, identical incident pressure).

So, for most of the spectrum, there will be a moderate increase in signal to noise ratio because of the 6 dB increase in pressure and the 3 dB increase in random noise, but as you ascend in frequency, phase differences will make the additional pressure less than 6 dB. Thus you can think of this as a noise reduction technique, but this buys you all of 3 dB of inproved signal to noise ratio (+6 dB signal + 3 dB noise = +3 dB signal), and that's not a lot. You would be far better off switching to a single omni with greater surface area (and having electronics at least as quiet as those found in the proposed mics) such as a 1" omni.

The rationale behind X-Y is (of course) that the cardiods' directivities (polar responses as a function of frequency) provide the left-to-right separation.

[Audiop]

Using two mic's this way will increase SNR 3dB (in the range below 10kHz or so where these mic's are essentially omnipolar) so yes, using two mic's makes sense if you have them and naturally there will be some differences between the channels since even the very omnidirectional EW omnis have some top roll off.

The recording will still be essentially mono though.

edit: Mark beat me! :-)

/Peter

[Audiop]

Originally Posted by Marshall Oliver

So with a with this technique, would it be right to say that would have the focus of a cardiod pattern, without the proximity effect?

Nope, not at all! :-)


/Peter

[Mark A. Jay]

Originally Posted by Audiop

Nope, not at all! :-)


/Peter

The cardiods' patterns govern the separation, and you won't get that with omni transducers in an X-Y configuration; Peter is 100% right - mathematically speaking, this is effectively mono. No more, no less.

[NorseHorse]

Omnis aren't omnis, folks.

Therefore, two omnis in XY will not be indentical to mono.

I don't necessarily recommend it to anyone who likes stereo though...

[jnorman]

peter - using two mics instead of one, why would you not also be increasing the self noise by 3dB?

[9sbean]

Originally Posted by davidvaldes

He claims that technique is described in the booklet accompaning the microphones...

Honestly, I´m really disconcerted, as I can´t see how omnis in XY can produce stereo.

No, I won't trust the booklets. Especially from the people that possibly did not practice recording at all.

Omnis are not really omni at HF, the worst the case, the better chance to get wider stereo.

So, if the omni did good stereo with XY setting, meaning they are not omni, or they are not good omni.

My 3 cents.

[Mark A. Jay]

Originally Posted by NorseHorse

Omnis aren't omnis, folks.

Therefore, two omnis in XY will not be indentical to mono.

I don't necessarily recommend it to anyone who likes stereo though...

Within limits and considering low frequency (especially), omni's are omni-directional; it is true to say that as frequency increases (and refraction effects are present) then they are no longer truly omni-directional. Were this not the case then every test lab on the planet would be guilty of generating data that violates fundamental principles of acoustics and measurement protocol.

Mind you, there are guidelines about when to use free-field, diffuse-field, and other types of mics (based on boundary conditions), but omni's... are omni's when used in observance with the specified guidlines and ISO standards, especially so for low frequency pressure perturbations (i.e. sound pressure).

You might wish to consult the people at companies such as Bruel & Kjaer, G.R.A.S. et al, or better, pick up a copy of "Acoustics" by Leo Beranek for more clarification.

As for using omni's in an X-Y configuration I stand by my assertion that mathematically, and especially for lower frequencies, the signals will be nearly-identical and thus, mono. I'm not sure either why aesthetically one would weant to do this, but the math supports my assertions.