Stereo Mic Technique

[midnightsun]

I am looking forward to getting some matched mics for stereo recording. I have search and have come up with bits and pieces of stereo micing techniques. I am interested in experimenting with all stereo micing approaches.

One approach that I am trying to understand is the MS technique. I understand that the M is a cardioid mic and the S comes from a perpendicular fig 8 mic with normal phase sent to the left phase reversed sent to right. I can sort of see that the stereo width comes from the proportion of M and S. Regarding the phase of M does it always sit with the left S or is it some how in between the phase of the S right and S left. I just don't quite get it. Or do you really need a MS encoder to make this approach work correctly?

[pkautzsch]

MS miking is:

(1) A mic of whatever pattern you choose pointing forward. Often it's a cardioid, but depending on what you like, you can also use an omni or hyper. Different patterns will result in different M/S ratios required for the same stereo width.

(2) A fig-8 mic with the positive lobe pointing to the left, the null pointing to front and back, and the negative lobe pointing to the right.

The L signal is the result of summing M and S mics, the L signal is the result of summing M and phase-flipped S. In a DAW you typically duplicate the S track and phase invert one. The non-inverted S is panned hard left, the inverted S is panned hard right. M is panned dead center.
In a mixing desk you might use the Direct Out or Aux of the S channel to duplicate it, send it to another input and flip that input's phase. Panning is the same.

[midnightsun]

Originally Posted by pkautzsch

MS miking is:

(1) A mic of whatever pattern you choose pointing forward. Often it's a cardioid, but depending on what you like, you can also use an omni or hyper. Different patterns will result in different M/S ratios required for the same stereo width.

(2) A fig-8 mic with the positive lobe pointing to the left, the null pointing to front and back, and the negative lobe pointing to the right.

The L signal is the result of summing M and S mics, the L signal is the result of summing M and phase-flipped S. In a DAW you typically duplicate the S track and phase invert one. The non-inverted S is panned hard left, the inverted S is panned hard right. M is panned dead center.
In a mixing desk you might use the Direct Out or Aux of the S channel to duplicate it, send it to another input and flip that input's phase. Panning is the same.

That clears up a great deal. I worked with the MS tracking approach and I like it. It is nice to get a stereo dimension without having to have match microphones. I think that this is a nice tool. Thanks for the clarity.

[Kiwiburger]

I remain skeptical about the use of M/S mic'ing. My reason: I like to ensure that my mixes sound best in mono, and get better in stereo. Old school, I know.

The problem (as I see it) with M/S is that the stereo side signal is created from a phase reversal of that figure 8 pointing left/right. So it's A/ not realistic, and B/ it's going to cancel out to zero when mono'd. Yes, zero. Complete nulling. Ambiance all gone - bye, bye.

If the mid mic is fairly dry, that means a drastic change in depth when mono'd. I think this is best used when the Mid mic is kept fairly ambiant, so this isn't such a rude shock when mono'd.

Just my 2 cents ... I prefer crossed or spaced pairs myself. They will have different issues when mono'd - if anything, the depth gets deeper when mono'd. Maybe there is good reason to mix these different stereo techniques up when layering BV's etc. When mono'd, some with get dryer, some will get wetter, and the end balance might be better preserved .... hmmm ...

[ine-kpro...]

Originally Posted by pkautzsch

MS miking is:

(1) A mic of whatever pattern you choose pointing forward. Often it's a cardioid, but depending on what you like, you can also use an omni or hyper. Different patterns will result in different M/S ratios required for the same stereo width.

(2) A fig-8 mic with the positive lobe pointing to the left, the null pointing to front and back, and the negative lobe pointing to the right.

The L signal is the result of summing M and S mics, the L signal is the result of summing M and phase-flipped S. In a DAW you typically duplicate the S track and phase invert one. The non-inverted S is panned hard left, the inverted S is panned hard right. M is panned dead center.
In a mixing desk you might use the Direct Out or Aux of the S channel to duplicate it, send it to another input and flip that input's phase. Panning is the same.

duplicate the S track and phaze invert one? can you phaze invert a recorded audio in your daw? how do you do that, is that a function in protools or logic? what is phaze. i have always known that when in stereo micing, you phaze invert one mic to prevent cancellation on stereo playback. exactly what is that?

[Kiwiburger]

Strictly speaking it's a polarity invert. But polarity invert is often called a phase switch - in the context of a balanced cable you are swapping the two phase wires around. You can do this by whatever means - many DAWs have a phase switch on each channel, or you can do an offline process.

I prefer to use the word 'phase' when describing a time offset, relative to another signal. This is not the case in this instance.

You are taking the mono signal from a figure eight mic, so basically the room reflections from left and right are mixed together into one. You are creating a fake stereo effect by multing/cloning this, inverting the polarity and panning these two signals left and right. It sounds stereo (just like stereo chorus pedals with only one delay but two outputs). But it's obviously not real, because the information from - say - the left hand side will appear on both sides. And they will 100% null to zero when mono'd.

If i'm doing this wrong, somebody please correct me.

[midnightsun]

Originally Posted by Kiwiburger

Strictly speaking it's a polarity invert. But polarity invert is often called a phase switch - in the context of a balanced cable you are swapping the two phase wires around. You can do this by whatever means - many DAWs have a phase switch on each channel, or you can do an offline process.

I prefer to use the word 'phase' when describing a time offset, relative to another signal. This is not the case in this instance.

You are taking the mono signal from a figure eight mic, so basically the room reflections from left and right are mixed together into one. You are creating a fake stereo effect by multing/cloning this, inverting the polarity and panning these two signals left and right. It sounds stereo (just like stereo chorus pedals with only one delay but two outputs). But it's obviously not real, because the information from - say - the left hand side will appear on both sides. And they will 100% null to zero when mono'd.

If i'm doing this wrong, somebody please correct me.

Kiwiburger-- you bring up excellent points that still baffle. Regarding the symantics-- The terms "phase, polatiry, and invert" are related but not the same. If a person is talking about what is taking place in air the term "phase" seems to be the only term that mathematically describes the gradations subtraction and addition that take place when actual sound waves interact with each other. Flipping a polarity switch or simply inverting the wave form is just saying that there is a 180 degree phase change. "Polarity or Invert" don't at all describe what is taking place when there is a time offset. With time offset phase changes from 0 to 360 degrees depending on the offset and specific frequency.

Kiwiburger-- you mention that when you split a signal and invert the polarity on one of the channels and listen mono that there is 100% null. I am certainly not going to argue against this point but wish to throw out the following observations and my own speculation-- True or pure sine waves only exist on paper and on some test equipment. Given that real acoustical sound waves are complex wave forms that require Fourier Transformation to accurately, mathematically describe. I am stuck having to conceptualize "Phase" issues with respect to sine waves because I am not up on Fourier Transformation. Given-- I just set up the MS mic configuration, recorded the fig 8 S track in a room with lots of ambience, split the S track and inverted the polarity of one of the tracks. I start by monitoring in stereo but only the Side tracks (+S left and -S right). I hear a very wet, pleasant "reverb" or ambience. I measure +S and -S on SpectraFoo Correlation Meter and it reads almost -1. When I hit the mono button on the analog console the sound almost disappears but not 100% gone. What remains is really, really wet. It may be that my audio pathways are flawed or something is going on here than I and many others don't fully understand. ???? If inverting polarity 100% nulls, then the downward component of the wave must be a perfect mirror image of the upward component. If the negative coordinates of the wave are not a 100% mirror image then during the subtraction process there will be something left over. Perhaps that is the faint very wet ambient sound than I heard as described above.

Hang in there I am almost done-- I go back to stereo monitoring and now bled in the M component of the MS setup. I like it because it sounds very natural and not like a reverb effect. I use SpectraFoo Correlation Meter to study the stereo mix. As one would expect when the side component is brought up the meter moves toward -1. It coincidentially turned out that the sound that was the best blend to my ear ended up with the correlation meter dancing between 0 and and half way toward +1.

Kiwiburger-- you mention that the MS stereo effect is artificial!!!! follow this... I kept my MS setup in place and moved the sound source (harmonica) significantly off center. When monitoring the MS data appropriately the stereo image accurately replicates the off center harmonica and sounded good. When I hit the mono button on the console the positional image disappeared but it still sounded good in mono. When I solo's the M channel it sounded dry. Solo'd the S channels they sounded very distant and wet.

Coming at this from another direction. In the mastering process one can use MS techniques to change the stereo image and width. Mid is a sum of right & left. Side is the difference of right & left (left added to polarity inverted right). If a stereo signal is encoded to MS and then decoded without any changes the final stereo signal will be the same as the original. Or, the MS channels can be seperately manipulated (compressed, EQ'd, or faded).

Is the figure 8 mic that is used to capture the Side signal capturing information specific to stereo image? My understanding is that it does????

[JGP]

MS stereo is completely mono compatible-when you sum it to mono the s and inverted s cancel out and the pure mono is left. No phaseyness at all. It is a very natural stereo sound and you can make it less stereo (and less ambient) by lowering the levels of the s
channels in your mix, or vice versa. I use a shure VP-88 which has the ms matrix in the mic and it is great as a room mic for live recording or as drum OH etc.

[d_fu]

Originally Posted by ine-kpro...

i have always known that when in stereo micing, you phaze invert one mic to prevent cancellation on stereo playback

I don't know where you got that from, but it's totally wrong...

[d_fu]

Originally Posted by Kiwiburger

You are taking the mono signal from a figure eight mic, so basically the room reflections from left and right are mixed together into one. You are creating a fake stereo effect by multing/cloning this, inverting the polarity and panning these two signals left and right. It sounds stereo (just like stereo chorus pedals with only one delay but two outputs). But it's obviously not real, because the information from - say - the left hand side will appear on both sides. And they will 100% null to zero when mono'd.
If i'm doing this wrong, somebody please correct me.

You are indeed wrong. An Fo8 microphone does have a direction. See pkautsch's posting above. The mic has a "positive" and a "negative" side. I came across a Wikipedia article the other day that explained it in easy terms, but I can't find it now. In any case, the stereo effect is not artificial.
You pan M to the middle, then, by panning one instance of S left, you add to the left channel (of the resulting stereo signal) what comes from the Fo8's positive/left side and subtract what comes from the negative lobe on the right. By reversing the S signal that goes to the right channel, you add what comes from the right (negative lobe, now turned to positive) and subtract what comes from the left of the Fo8 to the resulting stereo signal... Does this make sense?


Daniel

[midnightsun]

How does the Blumlein Technique compare to the MS? It is the same except that the M is a fig 8? It it endoced and decoded the same?

[pkautzsch]

Originally Posted by Kiwiburger

I
The problem (as I see it) with M/S is that the stereo side signal is created from a phase reversal of that figure 8 pointing left/right. So it's A/ not realistic, and B/ it's going to cancel out to zero when mono'd. Yes, zero. Complete nulling. Ambiance all gone - bye, bye.

Originally Posted by Kiwiburger

You are taking the mono signal from a figure eight mic, so basically the room reflections from left and right are mixed together into one. You are creating a fake stereo effect by multing/cloning this, inverting the polarity and panning these two signals left and right. It sounds stereo (just like stereo chorus pedals with only one delay but two outputs). But it's obviously not real, because the information from - say - the left hand side will appear on both sides. And they will 100% null to zero when mono'd.

You don't just hear the S signal in a decoded MS recording. What you hear is the SUM of M and S from the Left, and the DIFFERENCE of M and S from the right. It's not just using the same signal with one channel inverted.
When summing decoded MS to Mono, it's exactly the S signal cancelling out, and the M signal is left over.

Midnightsun:
Blumlein does not need to be decoded at all. It's just XY with another pattern. One mic goes hard left, one goes hard right. It can even be panned. You just have to make sure that NO sound source is at the sides of the setup (where the positive lobe of one mic and the negative lobe of the other cancel out the signal when summed to mono).

[Kiwiburger]

You don't just hear the S signal in a decoded MS recording. What you hear is the SUM of M and S from the Left, and the DIFFERENCE of M and S from the right. It's not just using the same signal with one channel inverted.
When summing decoded MS to Mono, it's exactly the S signal cancelling out, and the M signal is left over.

I understand - and that has been my point all along. When summed to mono, only the M signal is left - boosted. The fact that the encoded signal shares the Mid information doesn't negate the fact that the Side information is identical, just polarity reversed.

I think my problem with it is that it's not designed for close mic applications. My studio is too small for distant mic'ing techniques. I can imagine it could be very successful in larger spaces.

I still consider this to be a false stereo effect. A true stereo recording (if anything can be considered 'true' - perhaps a dummy head) would have slightly different reverb tail signals left & right, being the sound actually captured from the left and the right. They would NOT cancel out when mono'd. That only happens because the signals are in fact identical (a mixture of left & right) and their only point of difference being the polarity inversion.

For example: would you buy or use a digital reverb that simply inverted the phase of one side? I think not ... it would be a constant source of embarrassment when mono'd ....

Just my 2 cents ...

[cjogo]

Simple Stereo ----just correct your placement for phase--and the sound your are trying to achieve.
We generally use a Rode NT3 (hypercard) or a AT 4033 on the neck and a Rode NTK or U89 on the bottom

[midnightsun]

Originally Posted by pkautzsch

You don't just hear the S signal in a decoded MS recording. What you hear is the SUM of M and S from the Left, and the DIFFERENCE of M and S from the right. It's not just using the same signal with one channel inverted.
When summing decoded MS to Mono, it's exactly the S signal cancelling out, and the M signal is left over.

Thanks for the straight forward explaination. It makes total sense. I can't wait to put this to use as well as other stereo mic techniques. I have a great live room. It is almost 30x30x30 and has wood, assymetric walls with an atrium to carry away stray bass waves.

[Kiwiburger]

What you hear is the SUM of M and S from the Left, and the DIFFERENCE of M and S from the right. It's not just using the same signal with one channel inverted.

As I see it, it is exactly the same as far as the ambiant information from the Side mic is concerned.

Sum just means the Side mic is mixed with the Mid mic and panned Left.
Difference just means the Side mic is mixed with the Mid mic and panned right - but the polarity of the Side mic is inverted.

So far, everyone seems to agree that when mono'd, the Side information totally nulls, leaving the Mid information. That is telling you that the stereo effect is artificially created from a mono mic. It has some resemblance to reality only if an omni mic is used where the the patterns overlap and are picking up basically the identical waveforms, allowing real Sum & Difference effects.

But most pro's seem to endorse using a Cardioid for the Mid mic, defeating the basic principle if Sum & Difference. This gives a more pronounced difference between the two mic signals (because they have little in common), but IMO this leads to the disappearing fake stereo effect that I don't like.

I seem to be on my own here - even Wikipedia talks about using a cardioid for M/S micing. So i'll leave it at that - so far nobody has given me a compelling reason to use this technique over other stereo techniques.

[midnightsun]

Originally Posted by Kiwiburger

I still consider this to be a false stereo effect. A true stereo recording (if anything can be considered 'true' - perhaps a dummy head) would have slightly different reverb tail signals left & right, being the sound actually captured from the left and the right. They would NOT cancel out when mono'd. That only happens because the signals are in fact identical (a mixture of left & right) and their only point of difference being the polarity inversion.

For example: would you buy or use a digital reverb that simply inverted the phase of one side? I think not ... it would be a constant source of embarrassment when mono'd ....

Just my 2 cents ...

Kiwiburger... my developing understanding of MS is that it does capture "true stereo" information. I respectfully submit my following thoughts that make sense to me!

Imagine an MS setup in good position in a good room. 4 feet dead center is a singer and 45 degrees to the right is a mandolin. Two tracks are recorded Mid and Side. For all intents and purposes the performance has been “encoded” to MS. Now for stereo, decode the MS tracks by adding +S to M for left, and –S to M for right.

When monitoring the decoded MS as described above, without any M or S volume changes, the stereo image will be vocal dead center (equal volumes R & L), and the mandolin will be about 45 degrees to the right. If while encoded in MS the M is increased and S decreased, the decoded result will be that the mandolin is moved closer to the middle. If while encoded in MS the M in decreased and the S is increased, the decoded result will be that the mandolin is further to the right.

Why? It is intuitive that if M in increased that everything comes toward the middle. Less intuitive-- the S mic during the initial recording picked up 45 degrees more signal through the back side of the fig of 8 (the mandolin was on the right) and the M mic initial recorded the mandolin through the front side. Therefore, when adding the non polarity inverted S (+S) to M there will be phase cancellation of a portion of the mandolin. Thus, the left stereo field will have less mandolin. Again, the mandolin is mostly being picked up through the backside of the fig 8 S mic (i.e. there is a phase reversal). Therefore, if you invert the polarity of S (-S) and add it to M there is an additive summation of the mandolin. So if the entire S signal is increased there is more mandolin canceled from the left and more mandolin added to the right. (We would all accept that if we want to record through the back side of a fig of 8 mic all you have to do is invert the polarity to make things right.)

[Kiwiburger]

I agree - the theory of converting from L/R to M/S and back again is perfect, and I do play around with M/S processing on previously recorded tracks.

But for sum and difference to actually work, there needs to be a lot of overlap in the mic patterns.

If you have a close cardioid mic that rejects sound from the side (maybe my SM7), and a figure eight mic oriented left & right, strongly rejecting sound from the front and back ... they they don't have much in common.

I obviously don't have the space to do it right.

I can see that if you use a nice omni for Mid, and kept outside the nearfield zone, then M/S would give fairly good 'true stereo'.

Imagine a sound on the far left. Both the Mid and Side mic would pickup direct sound about the same. When encoded, the Mid and Side summed and panned to the left would get the boost of the two fairly identical signals. When the Mid is summed with the inverted Side mic and panned to the right, the two fairly identical signals will null.

When done properly, a good sense of the true left/right location would be apparant. And when mono'd, the Mid mic would still be fairly ambiant.

I can see the applications for people with far bigger studio space than me.

[kats]

I like MS used subtly, but anything more hurts my brain. It sounds more like an effect than anyhting natural.

The reason it's cool is that you still get a good solid sound when summed mono. Doing the same thing with effects can yield very unsatisfactory results in mono. And that's the point of MS IMO.

[d_fu]

Originally Posted by Kiwiburger

I understand - and that has been my point all along. When summed to mono, only the M signal is left - boosted.

No, it ain't boosted - why would it be?

The fact that the encoded signal shares the Mid information doesn't negate the fact that the Side information is identical, just polarity reversed.

Yes and no... Remember the fact that the polarity reversal turns the signal components from the Fo8's negative lobe into positive...

I think my problem with it is that it's not designed for close mic applications.

Says who?

For example: would you buy or use a digital reverb that simply inverted the phase of one side? I think not ... it would be a constant source of embarrassment when mono'd ....

Very funny... But irrelevant and based on misconception. M/S is true stereo, just like XY, which is just as mono-compatible.

Daniel

[pkautzsch]

Originally Posted by Kiwiburger

Sum just means the Side mic is mixed with the Mid mic and panned Left.
Difference just means the Side mic is mixed with the Mid mic and panned right - but the polarity of the Side mic is inverted.

I'll try to explain it a different way.
Imagine a cardioid being made up of an omni and a fig-8 with equal levels. By summing the different patterns, you get a pattern between it: Sound coming from the front comes out at the same level from both mics, thus getting boosted by 6 dB. Sound from 90° is not picked up by the fig-8, but fully by the omni, so it's only the omni level (6 dB lower than from front). Sound from the back is picked up by both mics at the same level, but one mic's signal is inverted, and thus they cancel out - resulting in NO signal.

It's similar with different angles between the mics. Using a cardioid center and a fig-8 side mic will result in a super-cardioid-ish pattern for LEFT, and inverting the fig-8 polarity results in a super-cardioid-ish pattern for RIGHT. You could do it with two fig-8 mics, one with the positive lobe to the left and panned left, one to the right panned right, and a center mic of any pattern. It DOES work. When it sounds phasey, something is wrong in the signal path or decoding.

[Kiwiburger]

Originally Posted by d_fu

No, it ain't boosted - why would it be?

erm ... because when your sum two identical signals you get a 6dB boost ... pkautzsch certainly understands this too ...

Yes and no... Remember the fact that the polarity reversal turns the signal components from the Fo8's negative lobe into positive...

Yes and yes ... L = M + S, R = M - S. Both left & right share the Mid information. The Side information (after encoding to stereo) is the same on both sides, just differing in polarity. This is the whole reason this side information nulls when monod. Or are you arguing that it doesn't null?

Says who?

Says me. I have used cardioids and figure eight on singer/guitar for the express purpose of seperation. This works extremely well, because you get very little guitar into the cardioid vocal mic, and very little vocal into the guitar mic. I could easily consider this to be a M/S arrangement, and encode it into stereo. I would then end up with a fake guitar track that disappears when mono'd - not what I consider satisfactory. This is my point: I don't believe using a close cardioid with good side rejection is a correct M/S technique. Are you saying it is?

Very funny... But irrelevant and based on misconception. M/S is true stereo, just like XY, which is just as mono-compatible.

XY (which I prefer by far) is not the same as M/S. X/Y is mono compatible in the sense that when summed the close direct sound gets a 6dB boost, and the ambiant information, being slightly different between left & right, gets about a 3dB boost. The ambiant information doesn't null and disappear.

I continue to believe that M/S micing only works 100% properly when using an omni for the mid mic, and retaining sufficient distance.

[pkautzsch]

Originally Posted by Kiwiburger

erm ... because when your sum two identical signals you get a 6dB boost ... pkautzsch certainly understands this too ...

When you sum two identical setups, you get 6 dB boost.
As the decoded L and R signals are not identical, but contain the S part in opposite polarity, this part will simply null out.

Yes and yes ... L = M + S, R = M - S. Both left & right share the Mid information. The Side information (after encoding to stereo) is the same on both sides, just differing in polarity. This is the whole reason this side information nulls when monod. Or are you arguing that it doesn't null?

The side information nulls. But the addition of positive or negative S signal to the same M signal results in two different signals.

I have used cardioids and figure eight on singer/guitar for the express purpose of seperation. This works extremely well, because you get very little guitar into the cardioid vocal mic, and very little vocal into the guitar mic. I could easily consider this to be a M/S arrangement, and encode it into stereo. I would then end up with a fake guitar track that disappears when mono'd - not what I consider satisfactory. This is my point: I don't believe using a close cardioid with good side rejection is a correct M/S technique. Are you saying it is?

You should not consider that setup an MS arrangement. You wouldn't call a setup with one close mic near the singer's mouth and another close mic near the guitar's soundhole a "spaced pair", would you.
MS, as spaced pairs and ORTF, is a main mic setup which has a certain angle that will appear fully spread between the speakers. The advantage of MS is that you can change that angle by changing the level of the S signal, whereas in spaced pairs you'd have to change the spacing.

XY (which I prefer by far) is not the same as M/S. X/Y is mono compatible in the sense that when summed the close direct sound gets a 6dB boost, and the ambiant information, being slightly different between left & right, gets about a 3dB boost. The ambiant information doesn't null and disappear.

Mono compatibility is not about boosting a signal. It's about not getting comb-filtering when summing a stereo signal to mono. As M contains all the sound and the addition of S will only add information about the direction we hear the source from, it is perfectly mono compatible. XY has the big disadvantage of the mics not pointing to the source and therefore coloring most of the direct sound.

I continue to believe that M/S micing only works 100% properly when using an omni for the mid mic, and retaining sufficient distance.

Retaining sufficient distance is essential for all main mic setups, and actually for all mikage. About the M pattern: it's the same math behind it, no matter what pattern. It's just different numbers.

[d_fu]

Originally Posted by Kiwiburger

erm ... because when your sum two identical signals you get a 6dB boost ... pkautzsch certainly understands this too ...

Yup, he understands they are not identical...

Or are you arguing that it doesn't null?

No. I'm just trying to convey that it is not "fake" stereo (as in artificial stereo "enhancement" of a monoish signal by way of some phase spread). If it were, an MS signal would have no real L/R localisation, which of course it does have.

I don't believe using a close cardioid with good side rejection is a correct M/S technique. Are you saying it is?

Sure. It's MS as per the very definition of MS (unless you use the mic in a spaced position).

XY (which I prefer by far) is not the same as M/S.

Mathematically, it is.

The ambiant information doesn't null and disappear.

Neither does it do with MS. The M channel also contains ambient signals.

I continue to believe that M/S micing only works 100% properly when using an omni for the mid mic, and retaining sufficient distance.

Define "100 % properly"... And define "sufficient distance". Sufficient for what?

[stuntbutt]

MS is not fake stereo. I do agree with Kiwi that MS with a cardioid sounds different than XY with cardiods. I know they are the same mathmatically but they don't sound the same. I also agree with Kiwi that when doing MS with a cardioid and you sum to mono, a lot of the ambience goes away. However, I LIKE this about MS. Have you ever noticed how summing something to mono makes it sound more ambient? The ear wants less verb in mono.

[andychamp]

Originally Posted by midnightsun

How does the Blumlein Technique compare to the MS? It is the same except that the M is a fig 8? It it endoced and decoded the same?

In a Blumlein setup you use two coincident fig.8 mics, set up very much as you would an XY pair.
They're just panned L/R, no special encoding.
The center of your stereo field is in the middle between the two mics.

Its mono compatibility isn't as perfect as with M/S, but Bruce Swedien uses that setup to record room ambience when re-amping dry synths, and feels it sounds very natural. Check the guest moderator archive for the Q&A session he did here.

[midnightsun]

Originally Posted by andychamp

In a Blumlein setup you use two coincident fig.8 mics, set up very much as you would an XY pair.
They're just panned L/R, no special encoding.
The center of your stereo field is in the middle between the two mics.

Its mono compatibility isn't as perfect as with M/S, but Bruce Swedien uses that setup to record room ambience when re-amping dry synths, and feels it sounds very natural. Check the guest moderator archive for the Q&A session he did here.

Thanks, I'll check out the Swedien archive. My personal approach is to always make sure that my mix work in mono. In the past I've used dual mics and blended to mono simply for tonal and ambient character, i.e. acoustic guitar. As I have recently been experimenting with the MS approach I am finding that it is a useful for sculpting tone in stereo but as others have pointed out the S disappears in pure mono. I am looking forward to getting my matched mic pair so I can experiment with stereo techniques to help with tone and ambience.

[aracu]

A very straightforward way of getting a great stereo recording is with
a stereo mic such as akg426 or sf24 recording in ms with a ms
monitor such as the ones built into the Sound Devices recorders.
There's no point in worrying about the physics of ms since it works
very well! If I needed to understand perfectly the physics behind
everything I used, I would end up living in a cave.

[soundbarnfool]

I've been experimenting recently with using a figure 8 for the M signal and moving the S mic around for better tone (i.e. actually closer to the source) while keeping the nulls perpendicular for phase issues. Anyone else trying this appproach? Almost a Blumlien-M/S hybrid (I am not using anything remotely like a matched pair though).

[aracu]

I experimented briefly with that without getting good results,
but it is worth trying. I also experimented with using two
sets of ms matched pairs (total of 4 mics or two stereo mics), one
close with both capsules set to fig 8's and one far pair with it's
mid mic set to omni. This seemed more promising but I would
need to try it in various rooms to get a better idea. Some people
use two mid cardioid mics instead of one (total of three mics) with
one of the mid cardioid mics pointing towards the back of the room,
especially for surround processing, sometimes refered to as "double ms".
I haven't tried it. Another thing to try is to use a condenser mic
as the mid mic and a ribbon mic with a fat low end as the side mic.