So you've got a stereo ribbon mic - how do you choose between Blumlein and Mid-Side?

[Corran]

I was just wondering about this. Since you've got the freedom to use either Blumlein or Mid-Side with a simple twist of 45 degrees, which do you choose? Does the room, or size of the group, or type of music make a difference to the choice?

I've gotten great results with both Blumlein and Mid-Side so I just was wondering what other people's thoughts were when using a single-point stereo ribbon, or even condenser mics.

On first thought, I was thinking MS would be better for louder music (like jazz) due to the S mic having the null towards the group. Blumlein for quieter classical stuff. But the room sound might also be a big determining factor. Blumlein for a good room, MS for not-so-good maybe?

[NorseHorse]

Thanks for posting this question. I was just thinking about it yesterday!

[John Willett]

Blumlein himself used mid/sides.

So record MS.

The advantage is that the mid mic. points forward - and - the unique thing about using two fig-8s in MS is that when you "steer" them to widen or narrow the image, the fig-8 pattern itself does not change shape.

[Corran]

Interesting. I didn't think about the polar pattern remaining unchanged in MS.

But isn't it a bad idea to use MS with a ribbon S mic when recording really soft music? I mean, if the null is towards already soft music it seems like the reflections are going to be whisper quiet, requiring ridiculous (or impossible) amounts of clean gain. Is this incorrect?

[boojum]

The MS hears "like" it is a coincident stereo mic. Part of its beauty is the ease with which the ratio of mid to side can be varied which changes your stereo width and will, if overdone, diminish you mid to nothing. So the thing about the ribbon just "hearing" the soft side sounds can somewhat mislead your thinking. The decoded effect is that of the coincident stereo mic.

[Corran]

Ah, that makes sense.

I'd still love to hear if people use each technique for different things, even if they are more-or-less the same thing.

[kstrauss]

Originally Posted by Corran

I'd still love to hear if people use each technique for different things, even if they are more-or-less the same thing.

I am not a fan of MS recording, but the one application I like it for is a solo instrument. The center image is very stable. I think the main advantage of MS is not so much the possibility of matrixing, rather, when using a cardiod for the M mic, the source is on-axis thereby avoiding off-axis coloration. Fig 8s generally do not exhibit much coloration off-axis so you might as well go Blumlein. One other point, normally with Blumlein I'll move the mics backwards and forwards to achive the desired stereo spread and/or ambience. With MS you might be tempted to favor the side, or mid mic rather than moving it, which is not the same thing.

[aracu]

Originally Posted by Corran

But isn't it a bad idea to use MS with a ribbon S mic when recording really soft music? I mean, if the null is towards already soft music it seems like the reflections are going to be whisper quiet, requiring ridiculous (or impossible) amounts of clean gain. Is this incorrect?

I haven't found there to be a noticable difference in volume level between MS and Blumlien in working with quiet sources. In a very noisy venue coincidant cardioids can work better though, with the angle narrower than Blumlien if necessary.

Blumlien is supposed to use two figure 8s, and MS is usually not used with a figure 8 mid, but it can be and will sound more "natural", sometimes cleaner, and closer to Blumlien.

With good stereo imaging software any coincident array can be adjusted as if it were a mid- side recording.

[summer_room]

In theory there should be no difference - but set the mic up, record in blumlein, then record in MS and listen to the results. A/B the mid signals solo.
Take the one that sounds best.

[Ozpeter]

In theory there should be no difference

I'm inclined to disagree.

A notable feature of a Blumlein configuration is that sound from the rear will be as well recorded as sounds from the front - but with left and right reversed. Either of these factors may be relevant in particular places or on particular occasions.

The amount of rear audio picked up by an MS pair is partly dependent on the M mic - whether hypercardioid at one extreme or omni at the other. So you have more choice over how the room reflections are treated, and you get more accuracy in their placement left/right.

[Hmmm, I'm overlooking the fact that the actual question relates to a pair of fig-of-8 mics in effect - well, the reversa of image bit is still true, and the fact that the M mic is pointing directly at the front and at the rear also makes a practical difference].

I would suggest making a simple pair of test recordings in MS and in Blumlein. Set the mics up in the middle of a room, and walk round the mic chanting the position you are in (front, left front, left rear, rear, right rear, right front, front again - or whatever). Maybe jangle some keys in your hand as you go. Compare and contrast the result, monitoring on speakers and then on headphones.

[David Spearritt]

This is getting off topic. The original question is if it is a stereo ribbon, ie fixed figure 8's, and therefore fixed Blumlein, which is better, pure Blumlein or MS Blumlein.

So please, no steering of patterns, no balancing of M and or S, ie no red herrings into the question. If you are recording MS Blumlein with ribbons and you start balancing M over S or vice versa then you will end up with no Blumlein and I suspect something pretty useless. Go back to cardioid M and be done with it. Your super accurate imaging (the whole point) just went out the window.

The answer to the original question is it depends.

Here are some pros and cons.
1. MS'ing is noisier than pure Blumlein as a matrix is involved.
2. Off axis is a moot point pretty much because the ribbons are so thin that the off axis response is the same from 0 to 45 and excellent at all freqs and will certainly not differ over the 45 deg of interest.
3. MS needs a matrix, pure can be into any preamp, simple.
4. MS is easier to align, because of clear geometric signals on the mics themselves.

In real terms there is almost no sonic difference, only ergonomics and perhaps a slightly noisier recording if the matrix is a bit average.

[d_fu]

Originally Posted by David Spearritt

1. MS'ing is noisier than pure Blumlein as a matrix is involved.
3. MS needs a matrix, pure can be into any preamp, simple.

How does the matrix introduce noise? It's just a simple polarity flip...
And of course you can use a simple preamp and do the matrix stuff later, even in your DAW software. Monitoring would require a matrix, in software or hrdware.

[Corran]

This is garnering some interesting discussions.

I am also confused why MS would induce noise. I personally usually matrix the pair manually with a Y split and polarity reversal adapter, so I can monitor it.

[David Spearritt]

Daniel, because the heart of the question is whether there is any advantage for a Blumlein final result of having the mic 0-90 or 45-45, then the matrixing afterwards in DAW or mixer is not relevant.

Even so, matrixing involves a sum and a difference of both signals and noise assuming it is done in the MS preamp.

[David Spearritt]

Originally Posted by Corran

I am also confused why MS would induce noise. I personally usually matrix the pair manually with a Y split and polarity reversal adapter, so I can monitor it.

So you are now mixing three signals instead of two, so more noise.

[d_fu]

Originally Posted by David Spearritt

Daniel, because the heart of the question is whether there is any advantage for a Blumlein final result of having the mic 0-90 or 45-45, then the matrixing afterwards in DAW or mixer is not relevant.

I'm afraid I'm not quite with you on that one... Please explain.

Even so, matrixing involves a sum and a difference of both signals and noise assuming it is done in the MS preamp.

I really don't think there is a potential noise issue here, unless the hardware were really crappy. Any preamp that will allow matricizing to a stereo L/R output, the matrix would be post-preamp, so where's the noise?

Originally Posted by David Spearritt

So you are now mixing three signals instead of two, so more noise.

Uhhh... No. You are mixing two signals on each channel... And what does the "instead of" refer to?

And if you only monitor L/R and record MS for later processing, there is no issue at all...

[Corran]

Even if you were to record all three channels at once (as I usually do), wouldn't the sum and difference cancel out the noise??

[d_fu]

Originally Posted by Corran

Even if you were to record all three channels at once (as I usually do), wouldn't the sum and difference cancel out the noise??

Depends on where you matricize. If you run three channels through AD separately, the two S channels' converter noise will not cancel out, because it's not identical... In theory, because in fact ambient noise or mic self noise will be higher.

Why record three channels anyhow?

[David Spearritt]

Because to do MS Matrixing with a simple mixer, you need channel 1 as M, channel 2 as +S and Channel 3 as -S, you then pan M to the middle, +S hard left and -S hard right. Three channels instead of two, more noise.

[Corran]

I do just for monitoring purposes. I guess I should later just duplicate the second track rather than use the recorded track - but I've just used what I had.

Anyway, I haven't actually used M-S in a while, so I will probably try some new things. I just got my B&O BM-5 in the mail, completely refurbished with new magnets/trafos/ribbons so that's why I posted this question.

[David Spearritt]

Originally Posted by d_fu

I'm afraid I'm not quite with you on that one... Please explain.

The comparison is between mic at 0-90 and 0-45, all other things being equal, ie we have to record L and R in both cases.

Not record L and R in one case and M and S in another and use post processing. If we allow post processing for one example, we do not have a comparison.

[Martin Kantola]

Originally Posted by David Spearritt

1. MS'ing is noisier than pure Blumlein as a matrix is involved.

The sensible thing is to do the matrix in digital, to avoid any issues with additional hardware.

Originally Posted by David Spearritt

2. Off axis is a moot point pretty much because the ribbons are so thin that the off axis response is the same from 0 to 45 and excellent at all freqs and will certainly not differ over the 45 deg of interest.

While it's true that a ribbon behaves well of axis, I wouldn't say that 45 versus 0 degrees doesn't matter. Especially if you have a soloist dead center.

Originally Posted by David Spearritt

3. MS needs a matrix, pure can be into any preamp, simple.

No, MS would go into two identical preamps just as Blumlein. Matrixing only after recording two tracks.

Originally Posted by David Spearritt

4. MS is easier to align, because of clear geometric signals on the mics themselves.

That's a point, someone should make a holder designed for Blumlein with markings. But with a stereo ribbon as in this case, it's not that difficult, only one angle to get right.

The important additional difference is to me the fact that MS allows for some minor adjustments in the mix, while Blumlein does not. So if your monitoring is not perfect while recording, it might be a good thing.

Martin

[David Spearritt]

Originally Posted by d_fu

I really don't think there is a potential noise issue here, unless the hardware were really crappy. Any preamp that will allow matricizing to a stereo L/R output, the matrix would be post-preamp, so where's the noise?

Its not the hardware that's crappy, it's the mic noise floor which gets amplified in the matrix.

M+S is 2 times the noise going to the L channel, M-S is two times the noise going to the right channel.

[David Spearritt]

I was trying to define differences, but they are small.

While it's true that a ribbon behaves well of axis, I wouldn't say that 45 versus 0 degrees doesn't matter. Especially if you have a soloist dead center.

Well I don't hear a difference with my SF24 in front of a soprano, comparing MS with pure Blumlein.

The important additional difference is to me the fact that MS allows for some minor adjustments in the mix, while Blumlein does not. So if your monitoring is not perfect while recording, it might be a good thing.

Martin

Here we go again.

[NorseHorse]

Can anyone else confirm this noise thing? I use MS Blumlein all the time, and haven't noticed anything. I always matrix in the box.

And has anyone clearly answered the original question yet?

Are there times when XY Blumlein is better than MS Blumlein? And how do you chose between the two?

Thanks!

[roonsbane]

I tend to use my SF24 in M/S the vast majority of the time. I like the solid center image, something that I don't hear if I use it in Bluemlein. I often use it as the front pair in a decca tree along with 4006's TLs. Then I will push the mid up till it sounds a bit monoish and then slightly back off. This setup gives a lot of flexibility depending on the acoustics when on location in an unknown room. I feel like I get a lot of reach into the orchestra and a lot of tonal flexibility as well between the chunky soft SF24 and the clear 4006's. I always matrix the M/S using a freeware Voxengo decoder plugin within Nuendo. I usually prep my final air mixes (when it's not live on air) in the studio with a real monitoring environment. I then have so much more flexibility being able to solo and tweak reverb returns and do some mastering to the finished product. I find it is very difficult to make tweaky adjustments with headphones, but I can get in the ball park.
Cameron

[fifthcircle]

Konrad hit on much of what I would have said... I'm not a huge fan of M-S, but I do find it has its uses.

While on paper a M-S pickup should be the same as an X-Y/Blumlein/etc... pickup (choose your pattern an insert technique), I find that in practice they sound VERY different.

For single point ensemble pickups I will use Blumlein every time. M-S can work really well for soloist and solo instrument pickups. I'll often go M-S for the piano in a concerto. i'll use it for solo classical guitar. It works great as a single point drum set pickup (especially with a hypercard in the center). It can also work really well in some orchestral recording situations as a main pair when you have flanks supporting it as the mono sum is perfect for broadcast (I've used omni in the center for this sometimes, fig-8 other times- just depends on if I am using woodwind mics).

The advantage of M-S is the on axis pickup of that center microphone. With the right pattern, it can really dig into the middle of an ensemble. The down side is if you are getting too much information in from the sides, you have serious phase issues.

Blumlein, on the other hand has a very precise left to right image due to the locations of the on-axis pickp and the locations of the other mic's nulls. This is IMO a great thing for small ensembles- especially string quartets. For orchestral stuff it works well, but with a flanking pair, you have two competing images. The tight and precise Blumlein and the more "diffuse" image of the spaced pair (flanks). Those two images have to be rectified and if you can do that, you've got the best of both worlds...

Anyways, a few thoughts...

--Ben

[David Spearritt]

Sigh. I give up.

[Martin Kantola]

Originally Posted by David Spearritt

Here we go again.

What are you talking about? Matrixing in the DAW is a great way to do it, what's your problem with that?

You're not only summing noise, also signal... (EDIT:and the noise is hopefully uncorrelated...)

Martin

[Martin Kantola]

Originally Posted by David Spearritt

The comparison is between mic at 0-90 and 0-45, all other things being equal, ie we have to record L and R in both cases.

Not record L and R in one case and M and S in another and use post processing. If we allow post processing for one example, we do not have a comparison

Say what? You want to use MS but don't allow a matrix? I'm very confused.

Martin