Sign in with
Sign up | Sign in
Your question

What Is "Mid-Side" Miking?

Last response: in Home Audio
Share
Anonymous
April 7, 2005 8:26:58 PM

Archived from groups: rec.audio.pro (More info?)

I once read an article that talked about recording a solo classical
guitar.

The engineer who wrote the article said that he used a "mid-side"
microphone arrangement to get the most natural stereo image from the
solo guitar. (I hope I am recalling the terminology correctly here.)

What is this technique, and what microphones are most recommended?

I am going to be recording a solo steel-string acoustic guitar, and I
want to get as much of an ambient feel as I can.

Many thanks in advance....

*andy*

More about : mid side miking

Anonymous
April 7, 2005 11:39:01 PM

Archived from groups: rec.audio.pro (More info?)

Have you tried Google?

EADGBE wrote:
> I once read an article that talked about recording a solo classical
> guitar.
>
> The engineer who wrote the article said that he used a "mid-side"
> microphone arrangement to get the most natural stereo image from the
> solo guitar. (I hope I am recalling the terminology correctly here.)
>
> What is this technique, and what microphones are most recommended?
>
> I am going to be recording a solo steel-string acoustic guitar, and I
> want to get as much of an ambient feel as I can.
>
> Many thanks in advance....
>
> *andy*
>
Related resources
Anonymous
April 7, 2005 11:56:45 PM

Archived from groups: rec.audio.pro (More info?)

EADGBE <hwbosshoss@aol.com> wrote:
>
>The engineer who wrote the article said that he used a "mid-side"
>microphone arrangement to get the most natural stereo image from the
>solo guitar. (I hope I am recalling the terminology correctly here.)
>
>What is this technique, and what microphones are most recommended?

This is a coincident pair where a figure-8 mike pointing to the side
is combined with a cardoid mike pointing straight ahead, and the two
are matrixed to produce right and left signals. It's effectively identical
to an X-Y pair except that the center of the stereo image is on-axis with
the mid mike, so it tends to minimize coloration in the middle of the field.

For a solo guitar, a good first try for this technique might be a
Beyer M160 and M130 pair.

Incidentally, M-S miking, as well as other stereo techniques, are
discussed extensively in the FAQ, and there is a nice tutorial on
stereophony on http://www.josephson.com.

Invariably for close-miked guitar you're going to be collapsing the
stereo image considerably, so you can't really consider it precisely
traditional stereophony.

>I am going to be recording a solo steel-string acoustic guitar, and I
>want to get as much of an ambient feel as I can.

In that case, your key is to get a good-sounding room and pull back using
whatever stereo technique you prefer to get some of the room sound. But
the key is the room. Pulling back in a room that sounds boxy and closetlike
just gives you a recording that is boxy and closetlike.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Anonymous
April 8, 2005 10:06:45 AM

Archived from groups: rec.audio.pro (More info?)

"EADGBE" <hwbosshoss@aol.com> wrote in message
news:1112916418.360249.226540@l41g2000cwc.googlegroups.com...
> I once read an article that talked about recording a solo classical
> guitar.
>
> The engineer who wrote the article said that he used a "mid-side"
> microphone arrangement to get the most natural stereo image from the
> solo guitar. (I hope I am recalling the terminology correctly here.)
>
> What is this technique, and what microphones are most recommended?
>
> I am going to be recording a solo steel-string acoustic guitar, and I
> want to get as much of an ambient feel as I can.

MS records using a "Mid" mike pointed forward (typically cardioid or
hypercardioid) and a "Side" mike pointed, logically enough, sideways
(figure-8). The positive-polarity of the side mike points left.

The signals are decoded by adding the M and S signal together, which
generates the left channel, and subtracting them, which generates the right
channel. (One way to subtract is to flip the polarity of the side signal,
then add to the mid.)

There are mixers and mike preamps specifically designed to do the decoding,
but another approach is to record the two mikes to their own channels and do
the decoding later; this lets you, for example, reduce the separation by
lowering the amount of side mike in the mix.

Ideally the microphones should be as similar to one another as possible. One
way is to use variable-pattern condenser mikes; another is to use the Beyer
M160 (hypercardioid) and M130 (figure-8) ribbon mikes that are designed with
very similar motors, specifically for use in MS setups. They are, however,
pretty low-output, so you'll need quiet preamps with lots of gain.

Peace,
Paul
Anonymous
April 8, 2005 10:11:54 AM

Archived from groups: rec.audio.pro (More info?)

Thanks so much for all of the help so far! But I am having trouble with
some of the terminology.

Scott said in his response that the two mics are "matrixed" together to
form a stereo image, and Paul said that the two mics are "decoded" to
form the stereo image.

Paul goes on to say this: "The signals are decoded by adding the M and
S signal together, which generates the left channel, and subtracting
them, which generates the right channel. (One way to subtract is to
flip the polarity of the side signal, then add to the mid.)"

What exactly does this mean? I would have thought it was as simple as
recording the two mics into separate channels and panning/EQ'ing from
there.

Do I need some special equipment, as was suggested by Paul?

Paul Stamler wrote:
> "EADGBE" <hwbosshoss@aol.com> wrote in message
> news:1112916418.360249.226540@l41g2000cwc.googlegroups.com...
> > I once read an article that talked about recording a solo classical
> > guitar.
> >
> > The engineer who wrote the article said that he used a "mid-side"
> > microphone arrangement to get the most natural stereo image from
the
> > solo guitar. (I hope I am recalling the terminology correctly
here.)
> >
> > What is this technique, and what microphones are most recommended?
> >
> > I am going to be recording a solo steel-string acoustic guitar, and
I
> > want to get as much of an ambient feel as I can.
>
> MS records using a "Mid" mike pointed forward (typically cardioid or
> hypercardioid) and a "Side" mike pointed, logically enough, sideways
> (figure-8). The positive-polarity of the side mike points left.
>
> The signals are decoded by adding the M and S signal together, which
> generates the left channel, and subtracting them, which generates the
right
> channel. (One way to subtract is to flip the polarity of the side
signal,
> then add to the mid.)
>
> There are mixers and mike preamps specifically designed to do the
decoding,
> but another approach is to record the two mikes to their own channels
and do
> the decoding later; this lets you, for example, reduce the separation
by
> lowering the amount of side mike in the mix.
>
> Ideally the microphones should be as similar to one another as
possible. One
> way is to use variable-pattern condenser mikes; another is to use the
Beyer
> M160 (hypercardioid) and M130 (figure-8) ribbon mikes that are
designed with
> very similar motors, specifically for use in MS setups. They are,
however,
> pretty low-output, so you'll need quiet preamps with lots of gain.
>
> Peace,
> Paul
Anonymous
April 8, 2005 10:21:16 AM

Archived from groups: rec.audio.pro (More info?)

"EADGBE" <hwbosshoss@aol.com> wrote in message
news:1112965914.405167.225000@g14g2000cwa.googlegroups.com...
>
> Thanks so much for all of the help so far! But I am having trouble with
> some of the terminology.
>
> Scott said in his response that the two mics are "matrixed" together to
> form a stereo image, and Paul said that the two mics are "decoded" to
> form the stereo image.
>
> Paul goes on to say this: "The signals are decoded by adding the M and
> S signal together, which generates the left channel, and subtracting
> them, which generates the right channel. (One way to subtract is to
> flip the polarity of the side signal, then add to the mid.)"
>
> What exactly does this mean? I would have thought it was as simple as
> recording the two mics into separate channels and panning/EQ'ing from
> there.
>
> Do I need some special equipment, as was suggested by Paul?
>
> Paul Stamler wrote:
>> "EADGBE" <hwbosshoss@aol.com> wrote in message
>> news:1112916418.360249.226540@l41g2000cwc.googlegroups.com...
>> > I once read an article that talked about recording a solo classical
>> > guitar.
>> >
>> > The engineer who wrote the article said that he used a "mid-side"
>> > microphone arrangement to get the most natural stereo image from
> the
>> > solo guitar. (I hope I am recalling the terminology correctly
> here.)
>> >
>> > What is this technique, and what microphones are most recommended?
>> >
>> > I am going to be recording a solo steel-string acoustic guitar, and
> I
>> > want to get as much of an ambient feel as I can.
>>
>> MS records using a "Mid" mike pointed forward (typically cardioid or
>> hypercardioid) and a "Side" mike pointed, logically enough, sideways
>> (figure-8). The positive-polarity of the side mike points left.
>>
>> The signals are decoded by adding the M and S signal together, which
>> generates the left channel, and subtracting them, which generates the
> right
>> channel. (One way to subtract is to flip the polarity of the side
> signal,
>> then add to the mid.)
>>
>> There are mixers and mike preamps specifically designed to do the
> decoding,
>> but another approach is to record the two mikes to their own channels
> and do
>> the decoding later; this lets you, for example, reduce the separation
> by
>> lowering the amount of side mike in the mix.
>>
>> Ideally the microphones should be as similar to one another as
> possible. One
>> way is to use variable-pattern condenser mikes; another is to use the
> Beyer
>> M160 (hypercardioid) and M130 (figure-8) ribbon mikes that are
> designed with
>> very similar motors, specifically for use in MS setups. They are,
> however,
>> pretty low-output, so you'll need quiet preamps with lots of gain.
>>
>> Peace,
>> Paul
>
Anonymous
April 8, 2005 11:17:59 AM

Archived from groups: rec.audio.pro (More info?)

"EADGBE" <hwbosshoss@aol.com> wrote in message
news:1112965914.405167.225000@g14g2000cwa.googlegroups.com...
>
> Thanks so much for all of the help so far! But I am having trouble with
> some of the terminology.

Typically a cardiod mic is pointed toward the center. This is the MID mic.
You can consider this to be the mono component.

A figure 8 mic is pointed to the sides. A figure 8 picks up very little
directly in front or in back or on the angles. It is a tight pattern that
blocks out everything except both far sides with 2 very narrow lobes 180
degrees apart. It looks just like a number 8 lying on its side. This is
the SIDE mic.

OK, now this is the part you probably don't understand. The figure 8 picks
up a 180 degree Left side and Right side, but left and right are out of
phase. When the right capsule is receiving positive pressure, the left side
is receiving negative pressure. OK, if you understand that, now we can do
some math.

A normal coincident stereo mic pattern is to place the mics close together
but pointing about 90 degrees spread apart. (a figure 8 is pointing 180
degrees apart, much wider than you'd ever want). Sometimes people like a
little wider spread than 90 or less often they like a little narrower
spread. It depends on what you are doing. In a concert hall you might want
a very wide spread. For a solo guitar you may want a narrow spread. What
is so cool about MS mic'ing is you have one mic that is straight ahead mono
and you have a mic that is spread apart farther than you'd ever want it to
be, so by mixing the two together, you can control how wide the spread is at
a later date when you can listen in your studio. You don't have to commit
to the amount of spread you end up with in advance. Pretty Cool, Huh?

Now the last part. The left ands right figure 8 information is out of
phase. And also, it outputs to your board as a single mic, not left and
right. So how do you generate a useful L and R signal from these 2 single
mics neither of which is a useable L or R signal? Well lets say the left
side of the figure 8 is positive when the right side is negative. If you
add the cardiod to the figure 8 in equal amounts, you get M+S = (center +
L - R). - R means the right channel is canceling out, so this combination
gives you a signal not straight ahead, and not all the way over to the side
to the side, but about 45 digress to the front left with the right channel
information canceling out. Hey, this would make a great left channel
signal! Just add the M and S together and you get LEFT! If you mix in more
cardioid it will be tighter to the front and if you mix in more figure 8 it
will be more spread out to the side. NOW - If you SUBTRACT the figure 8
from the cardiod you get M-S = (center + R - L). In this case the left
channel is canceling out and it gives you a signal pointed between front and
right depending on how much cardioid and figure 8 you combine also. This is
our Right channel which can be wider or narrower exactly the to exactly the
same spread as the Left assuming you keep the same gain structure. Just
subtract the M and S together and you get RIGHT! Just add the M and S
together and you get LEFT! An elegant flexible simple solution.

Don't let the matrix part throw you. Its just a complicated way of saying
there are several ways to combine the mics to get different image spreads.
The other important part to remember is we ADD to get one channel and
SUBTRACT to get the other. To do this you can just unsolder the wires to
that channel and reverse the polarity of the figure 8 signal. Let me also
add I over simplified the math a little because a cardiod mic actually
subtracts from the back in addition to picking up from the front. I should
have called the Mid mic = (center - some of the back too), but that makes
the math more confusing. The bottom line is that the back of a MS
combination mic doesn't pick up the back as well as the front and sides.

I have a little Sony ECM MS907 mic with 2 capsules built it. I use with my
pocket minidisc recorder. It has a 90 degree or 120 degree switch. It is
so useful that I can flick that one switch and change how tightly or how
loosely I an pinking up my image.

I hope that's clear. It either awful late or awful early right now. Maybe
Mark can say it more simply!

Julian
Anonymous
April 8, 2005 1:37:01 PM

Archived from groups: rec.audio.pro (More info?)

EADGBE wrote:
> Thanks so much for all of the help so far! But I am having trouble
> with some of the terminology.
>
> Scott said in his response that the two mics are "matrixed" together
> to form a stereo image, and Paul said that the two mics are
"decoded"
> to form the stereo image.

Different terms for the same basic process.

You've got two mics and two channels.

Both mics output information relating to the same two channels, but
neither mic has information relating to just one channel.

You've got to mix and match the mic outputs to get two signals that
you would recognize as the Left and Right channels. The device you use
to do this is called by some a matrix, and by others a decoder. It
might be made of resistors, transformers, op amps, a dedicated black
box, or even several whole channel strips on a mixing console. Or, it
might be some algebraic operations performed with DAW software.

> Paul goes on to say this: "The signals are decoded by adding the M
and
> S signal together, which generates the left channel, and subtracting
> them, which generates the right channel. (One way to subtract is to
> flip the polarity of the side signal, then add to the mid.)"

I'm quite sure that both Scott and Paul had exactly this in mind when
they wrote what they did.

> What exactly does this mean?

One man's matrix is another man's decoder. As they say, do the math.

> I would have thought it was as simple as
> recording the two mics into separate channels and panning/EQ'ing
from
> there.

If you have a console with polarity switches on the channel strips,
it's almost that easy.

> Do I need some special equipment, as was suggested by Paul?

Special is a state of mind. Us DAW guys can do it all in software.
However, left to my own devices I'd like to do it with a Rode NT-4
which is almost the same thing, only very different.
Anonymous
April 8, 2005 1:48:41 PM

Archived from groups: rec.audio.pro (More info?)

Arny Krueger <arnyk@hotpop.com> wrote:
>
>Special is a state of mind. Us DAW guys can do it all in software.
>However, left to my own devices I'd like to do it with a Rode NT-4
>which is almost the same thing, only very different.

Well, it's an X-Y pair anyway.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Anonymous
April 8, 2005 1:50:22 PM

Archived from groups: rec.audio.pro (More info?)

Scott Dorsey wrote:
> Arny Krueger <arnyk@hotpop.com> wrote:
>>
>> Special is a state of mind. Us DAW guys can do it all in software.
>> However, left to my own devices I'd like to do it with a Rode NT-4
>> which is almost the same thing, only very different.
>
> Well, it's an X-Y pair anyway.

Right, but X-y and M-S are both approaches to approximating intensity
stereo. ;-)
Anonymous
April 8, 2005 8:21:39 PM

Archived from groups: rec.audio.pro (More info?)

You take the sum and difference of the signals to form the left and right
channels. Thank about it.
Anonymous
April 8, 2005 8:53:48 PM

Archived from groups: rec.audio.pro (More info?)

On 4/8/05 10:17 AM, in article 115d4kpa5ruo85e@corp.supernews.com, "Julian
Adamaitis" <nospamJulianPA@Access4Less.net> wrote:
no component.
>
> A figure 8 mic is pointed to the sides. A figure 8 picks up very little
> directly in front or in back or on the angles. It is a tight pattern that
> blocks out everything except both far sides with 2 very narrow lobes 180
> degrees apart. It looks just like a number 8 lying on its side. This is
> the SIDE mic.

Nice basics on MS, especially on the hookup an dmatrix decoding ideas!
but for a couple points clarification:
Either I'm misreading you or a GOTTA disagree. Any decent bidirectional mic
is indeed 'DEAD-DEAD-DEAD' on EXACTLY the plane through the center, it's
called the NULL POINT cause it's...umm.. A POINT (or in this case a PLANE)
and anywhere BUT that point (or a point on that plane) it indeed hears
-something- starting IMMEDIATELY outside of that plane, it ramps up in
sensitivity pretty evenly and, barring bad design, it's freq response
on-axis isn;t too far off from even the extremes close to the null point.
A fig-8 mic with what you describe as 'very narrow lobes 180deg apart' isn;t
what you usually want for MS recording. The whole mathematical concept wants
a classic fig-8 pattern and depends on a good overlap of the MID mic and
SIDE mic.
Key to the MS approach working is that the mic elements must be positioned
ABSOLUTELY close together as physically possible, and they both must have as
wide and smooth a pattern throughout the spectrum as possible.

What
> is so cool about MS mic'ing is you have one mic that is straight ahead mono
> and you have a mic that is spread apart farther than you'd ever want it to
> be, so by mixing the two together, you can control how wide the spread is at
> a later date when you can listen in your studio. You don't have to commit
> to the amount of spread you end up with in advance. Pretty Cool, Huh?

Not quite. you;'re failing to keep separate the idea of PICKUP pattern and
directionality and PLAYBACK Sense Of 'width' or 'space. What you don;t
capture correctly with th emic rig, you can NOT make play back correctly
later.
Granted in toto that on PLAYBACK with 2 MS tracks to decode for stereo
playback, you control the apparent width of the final stereo soundstage by
going from M-ONLY to S-ONLY (with the caveat that S information decoded in
stereo is indeed still a mono signal, just reverse-polarity in the left vs
right speaker) and in decoded stereo playback (no matter WHAT the blend of M
vs S) -ALL- of the S-information cancels out and dissapears. Remember this
when you;re tempted to 'enhance' the decoded MS by over-accentuating the S
signal; Any sense of 'space' you over-enhance from decoding an MS recorded
pair into stereo with MORE S signal that normal, will VANISH in mono.

NEVER FORGET when you chose MS that your REAL sense of what's in front of
the mics comes from the -M- microphone. MS is wonderful with an OMNI there
but again, it's honest and you;d better like the room sound there. If you
MUST eliminate some part of the environment around the source and go to a
tighter pattern, be sure that your source can be seen completely by the M
mic or it won;t really be in the mix. This leads us back to PICKUP angle and
how that M mic is your -only- real look at the source, if the source is WIDE
relative to the mic pattern you;re in trouble, whether that's an average
not-too-wide cardioid mic 1 inch from an acoustic guitar (the guitar is WAY
wider than the mic pattern) or the same mic 20feet back from a ful symphony
and chorus (again, the extreme sides of the source are out of the mic's
clean pickup angle).
When decoded and played back in stereo such a recording will find that
whatever is at the extreme ends has gotten VERY indistinct as to WHERE it
seems to be and tends to vanish in mono. Anything that ONLY gets into the S
mic (cause it's outside the M mic pattern) will sound this way.

Wait'll we get into MS-Blumlein!
Anonymous
April 8, 2005 9:19:33 PM

Archived from groups: rec.audio.pro (More info?)

"EADGBE" <hwbosshoss@aol.com> wrote in message
news:1112965914.405167.225000@g14g2000cwa.googlegroups.com...
>
> Thanks so much for all of the help so far! But I am having trouble with
> some of the terminology.
>
> Scott said in his response that the two mics are "matrixed" together to
> form a stereo image, and Paul said that the two mics are "decoded" to
> form the stereo image.
>
> Paul goes on to say this: "The signals are decoded by adding the M and
> S signal together, which generates the left channel, and subtracting
> them, which generates the right channel. (One way to subtract is to
> flip the polarity of the side signal, then add to the mid.)"
>
> What exactly does this mean? I would have thought it was as simple as
> recording the two mics into separate channels and panning/EQ'ing from
> there.
>
> Do I need some special equipment, as was suggested by Paul?

Not if you have a computer and a multitrack program. Record the Mid and Side
signals onto two tracks. Make a duplicate track of the Side signal, then
invert its polarity. Pan the original Side signal to the left, the
cloned-and-inverted Side signal to the right, and the Mid signal straight up
the middle. Bingo -- you've just done the decoding/matrixing.

You'll probably find you want to lower the two Side signals by a certain
number of dB, depending on the "pan law" of your software. Here's how to
figure that out. Record or generate a tone (440Hz or whatever mid-frequency
suits your fancy) at, say, -5dBFS. Pan it hard left in your multitrack
program and set the fader so the output level meter reads -10dB. Pan it to
the center and see what each meter reads; if it's, say, 3dB lower than
hard-panned, then you'll need to lower the Side tracks by 3dB when you do MS
decoding. More if you want a narrower image, less if you want an exaggerated
image.

Peace,
Paul
Anonymous
April 8, 2005 10:31:21 PM

Archived from groups: rec.audio.pro (More info?)

SSJVCmag wrote:

> A fig-8 mic with what you describe as 'very narrow lobes 180deg apart' isn;t
> what you usually want for MS recording. The whole mathematical concept wants
> a classic fig-8 pattern and depends on a good overlap of the MID mic and
> SIDE mic.

The cardiod in the middle is really the superposition of a
mono and a figure 8 with the lobes pointing front and back
so that the directional components of the two mics are
orthogonal to each other (they don't overlap on each other's
main axis.) If they were mixed such that the side had half
the gain on its axis as the cardiod does on its axis then
the result of forming M+S and M-S is identical in theory to
two hypercardiods pointing at +/- 45 degrees. The angle
between and the front to back sensitivity ration of these
"logical" L and R mics can be varied by varying the mix of
the M and S, i.e. L = M + a*S and R = M - a*S.

> Key to the MS approach working is that the mic elements must be positioned
> ABSOLUTELY close together as physically possible, and they both must have as
> wide and smooth a pattern throughout the spectrum as possible.

Coincident on a vertical line is best with them as close as
possible along that line. It is less important that they be
wide and smooth than it is that they be the same. If the
frequency responses of the M and S are different, that
difference is manifest as a spatial spread in the acoustic
image of a single source having a wide spectrum output and
not directly on the front axis.

>
> What
>
>>is so cool about MS mic'ing is you have one mic that is straight ahead mono
>>and you have a mic that is spread apart farther than you'd ever want it to
>>be, so by mixing the two together, you can control how wide the spread is at
>>a later date when you can listen in your studio. You don't have to commit
>>to the amount of spread you end up with in advance. Pretty Cool, Huh?
>
>
> Not quite.

Pretty exactly, actually.

> Granted in toto that on PLAYBACK with 2 MS tracks to decode for stereo
> playback, you control the apparent width of the final stereo soundstage by
> going from M-ONLY to S-ONLY (with the caveat that S information decoded in
> stereo is indeed still a mono signal, just reverse-polarity in the left vs
> right speaker) and in decoded stereo playback (no matter WHAT the blend of M
> vs S) -ALL- of the S-information cancels out and dissapears. Remember this
> when you;re tempted to 'enhance' the decoded MS by over-accentuating the S
> signal; Any sense of 'space' you over-enhance from decoding an MS recorded
> pair into stereo with MORE S signal that normal, will VANISH in mono.

This is incorrect pretty much in toto.

>
> NEVER FORGET when you chose MS that your REAL sense of what's in front of
> the mics comes from the -M- microphone.

Anything exactly in front will only stimulate the M mic so
that you end up with it as a mono signal on both the L and R.

> MS is wonderful with an OMNI

What that gives you is logical L and R mics which are two
...cardiods pointing exactly left and exactly R. The amount
of M mixed in controls the pattern from subcardiod to
hypercardiod but the orientation remains side to side if M
is omni. Not many people would consider that M/S.


> be sure that your source can be seen completely by the M
> mic or it won;t really be in the mix.

Incorrect. If it's wide, its width will be in the
reproduction. I've done guitar where the fingering noises
were predominant in one channel and the rhythmic body
slapping were in the other with the picking sounds in the
center. Is this bad? I dunno.


Bob
--

"Things should be described as simply as possible, but no
simpler."

A. Einstein
Anonymous
April 9, 2005 5:49:05 AM

Archived from groups: rec.audio.pro (More info?)

On Fri, 08 Apr 2005 16:53:48 GMT, SSJVCmag <ten@nozirev.gamnocssj.com>
wrote:

>> A figure 8 mic is pointed to the sides. A figure 8 picks up very little
>> directly in front or in back or on the angles. It is a tight pattern that
>> blocks out everything except both far sides with 2 very narrow lobes 180
>> degrees apart. It looks just like a number 8 lying on its side. This is
>> the SIDE mic.
>
>Nice basics on MS, especially on the hookup and matrix decoding ideas!

Indeed. Welcome to the several new posters. Where have y'all been?

Chris Hornbeck
6x9=42
Anonymous
April 9, 2005 3:24:34 PM

Archived from groups: rec.audio.pro (More info?)

One of the best description of MS miking is on the back pages of the
Sennheiser MKH microphone brochure.

If you want to proceed further, there is an article by Manfred Hibbing in
the Oct 89 issue of the Journal of Audio Engineering Society comparing XY
and MS technicques. Further back, in the Oct 82 edition of the same
publication, Dooley and Streicher wrote a fairly complete discussion of MS.

Bruce Bartlett, of Shure Bros., wrote a good article on general stereo mike
technicques back in the Oct 79 issue of DB magazine.

Norm Strong
Anonymous
April 9, 2005 4:51:44 PM

Archived from groups: rec.audio.pro (More info?)

On 4/8/05 9:31 PM, in article d37b9c02fnv@enews2.newsguy.com, "Bob Cain"
<arcane@arcanemethods.com> wrote:

>> SSJVCmag wrote:
>>
>> A fig-8 mic with what you describe as 'very narrow lobes 180deg apart' isn;t
>> what you usually want for MS recording. The whole mathematical concept wants
>> a classic fig-8 pattern and depends on a good overlap of the MID mic and
>> SIDE mic.
>
> The cardiod in the middle is really the superposition of a
> mono and a figure 8 with the lobes pointing front and back

Wha? The bi-mic's lobes are not oriented front-to-back...


> so that the directional components of the two mics are
> orthogonal to each other (they don't overlap on each other's
> main axis.)

With an omni in MID they indeed DO overlap.
With a card (a proper card) they do as well, tho at the extremes the card's
down a bit but not GONE.
If they don;t overlap, there IS nothing to add/subtract and derive a stereo
field from... You have an odd single CENTER signal with this weird
rev-polarity ambience field.


> If they were mixed such that the side had half
> the gain on its axis as the cardiod does on its axis then
> the result of forming M+S and M-S is identical in theory to
> two hypercardiods pointing at +/- 45 degrees. The angle
> between and the front to back sensitivity ration of these
> "logical" L and R mics can be varied by varying the mix of
> the M and S, i.e. L = M + a*S and R = M - a*S.

Yep.


>
>> Key to the MS approach working is that the mic elements must be positioned
>> ABSOLUTELY close together as physically possible, and they both must have as
>> wide and smooth a pattern throughout the spectrum as possible.
>
> It is less important that they be
> wide and smooth than it is that they be the same.
> if the frequency responses of the M and S are different, that
> difference is manifest as a spatial spread in the acoustic
> image of a single source having a wide spectrum output and
> not directly on the front axis.

And the same thing/different flavor results if either mic has an overly
narrow pattern. Either deficiency results in odd/bad stereo.


>>> What
>>> is so cool about MS mic'ing is you have one mic that is straight ahead mono
>>> and you have a mic that is spread apart farther than you'd ever want it to
>>> be, so by mixing the two together, you can control how wide the spread is at
>>> a later date when you can listen in your studio. You don't have to commit
>>> to the amount of spread you end up with in advance. Pretty Cool, Huh?
>>
>>
>> Not quite.
>
> Pretty exactly, actually.

Nope, You can get -real- 'spread' only as far as the MID mic has a good
pattern. You can then indeed control your spread as you REDUCE the spread
from that to mono. Any further pushing of the SIDE element to invent 'wider'
results in a faux indistinct stereo ambience that does Interesting-But-Odd
things to the signal... It's a fun gimmick but it means that in mono the
level drops to whatever the MID mic is dialed in to, and just as a special
fun case, fed through a Dolby Stereo surround decoder, most of it goes to
the surrounds!


>
>> Granted in toto that on PLAYBACK with 2 MS tracks to decode for stereo
>> playback, you control the apparent width of the final stereo soundstage by
>> going from M-ONLY to S-ONLY (with the caveat that S information decoded in
>> stereo is indeed still a mono signal, just reverse-polarity in the left vs
>> right speaker) and in decoded stereo playback (no matter WHAT the blend of M
>> vs S) -ALL- of the S-information cancels out and dissapears. Remember this
>> when you;re tempted to 'enhance' the decoded MS by over-accentuating the S
>> signal; Any sense of 'space' you over-enhance from decoding an MS recorded
>> pair into stereo with MORE S signal that normal, will VANISH in mono.
>
> This is incorrect pretty much in toto.

Please, somebody, show me either where I've somehow either COMPLETELY
managed to obliviously miswrite myself in this thread after two sets of
proofing passes (it's happened...), or where 30 years of study and practice
with this in classical and field and post work is in the dumper...?
Somebody's in the wrong ballpark here.


>> NEVER FORGET when you chose MS that your REAL sense of what's in front of
>> the mics comes from the -M- microphone.
>
> Anything exactly in front will only stimulate the M mic so
> that you end up with it as a mono signal on both the L and R.

That 'exactly' is key. It's the singular null point for the bidirectional's
pattern. Move a degree or 2 left or right and you;re getting more and less
into both mics, and that;s why the listener in the resultant stereo playback
indeed CAN follow those moves and positionings so nicely.


>
>> MS is wonderful with an OMNI
>
> What that gives you is logical L and R mics which are two
> ..cardiods pointing exactly left and exactly R. The amount
> of M mixed in controls the pattern from subcardiod to
> hypercardiod but the orientation remains side to side if M
> is omni. Not many people would consider that M/S.

This 'not many people" would be the set including only engineers and
recordists who understand what MS is all about, and your classification
would remove a great number of marvelous classical recordings from the MS
roster.

Again; you pick your preferred recording POV by choosing your MID mic.
If you;d hang an omni there, you use an OMNI as mid, IF you need to
eliminate the environment to some degree, you move into some flavor of
cardioid. Once you're moving on past classic card into something narrower,
you're starting to toss parts of the baby out with the bathwater (until
you're all the way over to using two bidirectionals but then you're doing
MS-Blumlien which is a whole other wonderful thing...with its own special
case limitations). There may well BE a damned-but-compelling reason that you
are willing to reduce good imaging for a tighter pattern (usually it's to
desperately try and reduce some form of distracting ambient noise like fans,
crickets or candy wrappers) but you need to be VERY careful about knowing
how you're progressively eliminating the real interaction of both patterns
that makes the clean accurate stereo playback happen at all.


>> be sure that your source can be seen completely by the M
>> mic or it won;t really be in the mix.
>
> Incorrect. If it's wide, its width will be in the
> reproduction.

But at this point, a VERY significant part is NOT real stereo (real L/R
imaging information). It's probably INTERESTING, you may well LIKE it on the
studio monitors and thus It may well be USEFUL, just as long as you
understand what's going to happen to that effect when it's played back in
mono, or after going through any number of mangling transmission changes
like FM broadcast, mono playback or MP3 conversion.


> I've done guitar where the fingering noises
> were predominant in one channel and the rhythmic body
> slapping were in the other with the picking sounds in the
> center. Is this bad? I dunno.

How'd it sound checked in mono? In stereo headphones?
It runs into the same issue as micing an openback guit cab front and rear
and failing to rev the polarity on the rear mic. When panned hard L/R it
sounds 'interesting' and whacky-wide, but mono-check it (or listen from a
far enough distance where they combine in the room) and you hear the big
prob. John Hiatt's recording of IT FEELS LIKE RAIN is a depressing example
of this.


Again, I'm holding to everything I've blabbed here.
No changes. Only thing might be you're using most all of the terms WHOLLY
different/opposite than anyone I've ever chatted about this material with
and we need to hash that out.

I'm out for coffee...
John
Anonymous
April 9, 2005 4:51:45 PM

Archived from groups: rec.audio.pro (More info?)

>> The cardiod in the middle is really the superposition of a
>> mono and a figure 8 with the lobes pointing front and back

> Wha? The bi-mic's lobes are not oriented front-to-back...

You misread.

The easiest way to understand M-S miking is to assume the middle mic is omni.
Adding and subtracting the figure-8 signal produces two cardioid patterns
pointing left and right.

If the middle mic is cardioid, its figure-8 components rotates the virtual
cardioid patterns forward.
Anonymous
April 9, 2005 5:26:35 PM

Archived from groups: rec.audio.pro (More info?)

On 4/9/05 8:51 AM, in article BE7D4820.530A%ten@nozirev.gamnocssj.com,
"SSJVCmag" <ten@nozirev.gamnocssj.com> wrote:

> On 4/8/05 9:31 PM, in article d37b9c02fnv@enews2.newsguy.com, "Bob Cain"
> <arcane@arcanemethods.com> wrote:
>
>>> SSJVCmag wrote:
>>>
>>> A fig-8 mic with what you describe as 'very narrow lobes 180deg apart' isn;t
>>> what you usually want for MS recording. The whole mathematical concept wants
>>> a classic fig-8 pattern and depends on a good overlap of the MID mic and
>>> SIDE mic.
>>
>> The cardiod in the middle is really the superposition of a
>> mono and a figure 8 with the lobes pointing front and back
>
> Wha? The bi-mic's lobes are not oriented front-to-back...

Ah, were you describing 'how to derive a cardioid from an omni-plus-fig8
pattern"? If so you confused me by mistyping 'MONO' for 'OMNI' and I thought
you were talking about how to allign an MS rig. Still not sure what you were
getting at there though.


>> so that the directional components of the two mics are
>> orthogonal to each other (they don't overlap on each other's
>> main axis.)

Except if you WERE describing how to derive a card, inded the two pattersn
would be EXACTLY alligned along their main axes... I'm guess I'm still
really confused.
April 10, 2005 8:46:29 PM

Archived from groups: rec.audio.pro (More info?)

On 8 Apr 2005 06:11:54 -0700, "EADGBE" <hwbosshoss@aol.com> wrote:

>What exactly does this mean? I would have thought it was as simple as
>recording the two mics into separate channels and panning/EQ'ing from
>there.
>
>Do I need some special equipment, as was suggested by Paul?

Yes you need a MS decoder If you use Waves software there is one in
the bundle. The decoder or matrix does the math for you adding the two
signals and subtracting inversions of the mid from opposite sides to
create a stereo image. It's like using a calculator, you don't need to
know how to add<G>
There are of course hardware boxes that will do the same.
Anonymous
April 10, 2005 9:06:32 PM

Archived from groups: rec.audio.pro (More info?)

On 4/10/05 12:46 PM, in article sqli51likb7mbo91r7bq4q68h8eqbfgnak@4ax.com,
"paul@nospam.net" <paul@nospam.net> wrote:

> On 8 Apr 2005 06:11:54 -0700, "EADGBE" <hwbosshoss@aol.com> wrote:
>
>> What exactly does this mean? I would have thought it was as simple as
>> recording the two mics into separate channels and panning/EQ'ing from
>> there.
>>
>> Do I need some special equipment, as was suggested by Paul?
>
> Yes you need a MS decoder If you use Waves software there is one in
> the bundle. The decoder or matrix does the math for you adding the two
> signals and subtracting inversions of the mid from opposite sides to
> create a stereo image. It's like using a calculator, you don't need to
> know how to add<G>
> There are of course hardware boxes that will do the same.

The previous instructions also work real well...
Short n sweet:

Dedicate 3 tracks
Label them in order:
S-, M S+

Take the original SIDE mic track and make it S+ channel
COPY that track over to the S- channel
INVERT (polarity) the S- track
Put the MID track in the M channel

ABSOLUTELY match the levels of the S- and S+ tracks

(without the S-/S+ tracks being EXACTLY THE SAME at all times,
MS decoding doesn;t work.. I can only cringe at the possibilites in a DAW
system that might inadvertantly introduce a sample or 6 difference between
any of these 3 tracks by having one little THING in the system set
differently...)

Oversimply, varying the relative levels of
S-tracks vs M
changes your resultant L/R stereo image.


COPY it to another
Anonymous
April 11, 2005 4:45:42 AM

Archived from groups: rec.audio.pro (More info?)

On Sun, 10 Apr 2005 16:46:29 GMT, paul@nospam.net wrote:

>On 8 Apr 2005 06:11:54 -0700, "EADGBE" <hwbosshoss@aol.com> wrote:
>
>>What exactly does this mean? I would have thought it was as simple as
>>recording the two mics into separate channels and panning/EQ'ing from
>>there.
>>
>>Do I need some special equipment, as was suggested by Paul?
>
>Yes you need a MS decoder

No you don't, if you have three tracks free it's very easy to do it
with the mixer... even easier if you are doing it in a software mixer
in the computer. The Waves thing is handy but not essential.

Al

If you use Waves software there is one in
>the bundle. The decoder or matrix does the math for you adding the two
>signals and subtracting inversions of the mid from opposite sides to
>create a stereo image. It's like using a calculator, you don't need to
>know how to add<G>
>There are of course hardware boxes that will do the same.
Anonymous
April 14, 2005 10:52:10 PM

Archived from groups: rec.audio.pro (More info?)

SSJVCmag wrote:
> On 4/8/05 9:31 PM, in article d37b9c02fnv@enews2.newsguy.com, "Bob Cain"
> <arcane@arcanemethods.com> wrote:
>
>
>>>SSJVCmag wrote:
>>>
>>>A fig-8 mic with what you describe as 'very narrow lobes 180deg apart' isn;t
>>>what you usually want for MS recording. The whole mathematical concept wants
>>>a classic fig-8 pattern and depends on a good overlap of the MID mic and
>>>SIDE mic.
>>
>>The cardiod in the middle is really the superposition of a
>>mono and a figure 8 with the lobes pointing front and back
>
>
> Wha? The bi-mic's lobes are not oriented front-to-back...

Sorry for the mono-for-omni substitution. Damn four letter
words. :-)

In an MS we conceptually have three elements, a side to side
fig 8, a front to back fig 8 and an omni. Two of them, the
omni and front to back fig 8, are summed in the M signal.
If the omni and fig 8 components in the M have the same
sensitivity it is a cardiod.

For sound coming at the M axis from an angle theta (positive
to the right), a 1st order mid has the sensitivity

M = a + b*cos(theta)

where a and b are the gains of the M components which
determine the type of the M mic according to

a = b -> cardiod
a = 0 -> fig 8
a < b -> hypercardiod
a > b -> sub cardiod
b = 0 -> omni.

The fig 8 side sensitivity is

S = -c*sin(theta)

where c is the gain of the S.

Thus the sensitivity of L after mixing is

L = M + S = a + b*cos(theta) - c*sin(theta)

and that of R after mixing is

R = M - S = a + b*cos(theta) + c*sin(theta).

With some trig these become,

L = a + sqrt(b^2 + c^2)*cos(theta + atan(c/b))

and

R = a + sqrt(b^2 + c^2)*cos(theta - atan(c/b))

which are of the form of two logical 1st order mics
forming equal angles on either side of the M axis.

a, b and c determine the type of the logical mics according to

a = sqrt(b^2 + c^2) -> cardiod
a = 0 -> fig 8
a < sqrt(b^2 + c^2) -> hypercardiod
a > sqrt(b^2 + c^2) -> sub cardiod

The angle the logical L and R mics make to the front M axis
is atan(c/b).

If a, b and c are equal as in the case of a cardiod M and an
S with half the on axis gain of the M then the logical L and
R are at an angle of +/- 45 degrees and the pattern is
hypercardiod by a factor of 1.414 (the square root of two.)

If b = 0 (omni mid) then L and R are cardiods pointing due
left and due right. If c = 0 (no S in the mix) then L and R
are both a front pointing cardiod. If a = 0 then we have a
Blumlein configuration.

Now, all that is well and good except that it assumes that
the omni and fig 8 components of each mic are not frequency
dependant. Closer to reality we have

M = a(w) + b(w)*cos(theta)

and

S = -c(w)*sin(theta)

Which says that a, b and c, both magnitude and phase, are a
function of frequency rather than constant. This means that
the logical L and R sensitivity, the angle between them and
their pattern are all a function of frequency.

Only if b(w)/c(w) is a constant, i.e. well matched, will the
angle between the logical L and R be independant of
frequency. This means that unless M and S are well matched
in their directional components, a broad band source will be
smeared over a range of angles in the recorded image
depending on frequency. A drum set physically to the side
can sound as if the kick is at the side and the high hat is
in the center even when the set is a good distance from the
mic. Worse, if the phases of a, b and c are not the same,
and they won't be in general, the angles which the logical
mics make to each other are actually waggling in time in a
spatial Lissajou pattern.

This is why I like using two back to back cardiod capsules,
well matched to the cardiod M, subtracted from each other to
create a fig 8 S instead of using different types of mics
for M and S. If the three capsules used in this
configuration are well matched, the angle between the
logical L and R mics is independant of frequency. The
sensitivity (usually called frequency response) and their
pattern will vary in a way very similar to that of the
capsules used.


Bob
--

"Things should be described as simply as possible, but no
simpler."

A. Einstein
Anonymous
April 15, 2005 7:31:51 AM

Archived from groups: rec.audio.pro (More info?)

On 4/14/05 9:52 PM, in article d3n6o608ca@enews1.newsguy.com, "Bob Cain"
<arcane@arcanemethods.com> wrote:

>>> The cardiod in the middle is really the superposition of a
>>> mono and a figure 8 with the lobes pointing front and back
>>
>>
>> Wha? The bi-mic's lobes are not oriented front-to-back...
>
> Sorry for the mono-for-omni substitution. Damn four letter
> words. :-)

Really!

>
> In an MS we conceptually have three elements,
....
> M = a + b*cos(theta)
(SNIP THE PHOTOCOPY FROM THE THEORY BOOK...)
> This is why I like using two back to back cardiod capsules,
> well matched to the cardiod M, subtracted from each other to
> create a fig 8 S instead of using different types of mics
> for M and S. If the three capsules used in this
> configuration are well matched, the angle between the
> logical L and R mics is independant of frequency. The
> sensitivity (usually called frequency response) and their
> pattern will vary in a way very similar to that of the
> capsules used.

My Ghod Bob... What happened to
"Things should be described as simply as possible, but no
simpler." ???

All that just to say that IDEALLY one should have all-dead-on-matched mic
elements in a coincidental array, and so you do.
Whew...

What bothers me is the trade-off between the reality of trying to mount 3
or 4 cardioid mics that way and the inherent distances (in centimeters!)
between the diaphragms what SHOULD be -zero- and COULD be closer by at least
an order of magnitude, and what that does to your resultant pattern
combinations... Aren;t you better off with well-chosen mics that work
together well (as has been the case for over 1/2 century) with a
double-diaphragm bi-mic or good ribbon, than with the virtual bidirectional
element having a width measurable in inches?
This assumes of course that one doesn;t have a special design mic system
like the Soundfield handy...
Or did I miss the lessaon (and yes I DID read through it all!)
Anonymous
April 15, 2005 7:31:52 AM

Archived from groups: rec.audio.pro (More info?)

SSJVCmag wrote:

>>In an MS we conceptually have three elements,
>
> ...
>
>> M = a + b*cos(theta)
>
> (SNIP THE PHOTOCOPY FROM THE THEORY BOOK...)

Hey! It wasn't necessasary to open a book to write that.

>
>>This is why I like using two back to back cardiod capsules,
>>well matched to the cardiod M, subtracted from each other to
>>create a fig 8 S instead of using different types of mics
>>for M and S. If the three capsules used in this
>>configuration are well matched, the angle between the
>>logical L and R mics is independant of frequency. The
>>sensitivity (usually called frequency response) and their
>>pattern will vary in a way very similar to that of the
>>capsules used.
>
>
> My Ghod Bob... What happened to
> "Things should be described as simply as possible, but no
> simpler." ???

That's about as simple as it can be said and still be
reasonably accurate and predictive. I could have used
complex vector algebra. :-)

> All that just to say that IDEALLY one should have all-dead-on-matched mic
> elements in a coincidental array, and so you do.
> Whew...

Right. I'm just trying to show exactly what the effects of
them not being matched are. It is not all that well known
that the logical angle formed by L and R is dependant on the
matching and I'm not sure anyone has noticed the waggle
effect that comes from phase differences.

> What bothers me is the trade-off between the reality of trying to mount 3
> or 4 cardioid mics that way and the inherent distances (in centimeters!)
> between the diaphragms what SHOULD be -zero- and COULD be closer by at least
> an order of magnitude, and what that does to your resultant pattern
> combinations...

As long as they are coincident on a vertical axis they are
coincident for planar recording. The angles of sources off
the horizontal plane in real situations in which I use them
are not that great and can be pretty much ignored. The
things I've built and really like have a vertical stack of
SMS with the two S capsules pointing at opposite sides. SSM
probably can't be distinguished from that for the same
reasons that it works to begin with and I haven't tried it.

> Aren;t you better off with well-chosen mics that work
> together well (as has been the case for over 1/2 century) with a
> double-diaphragm bi-mic or good ribbon, than with the virtual bidirectional
> element having a width measurable in inches?

The double diaphragm bi is a really good way to get a
matched situation if it is used for both and they are well
matched at their fig 8 setting. A ribbon and a condensor
will have the problems with a frequency dependant angle that
I described.

I'll bet that if you set up an experiment with a compact
drum set say 15 feet away at an angle of 50 degrees or so
using a condenser M and a ribbon S with the usual HF rolloff
that you will hear what I described. Listening to the
recording, the lower mids will be about where they are
supposed to be and the cymbals will be much nearer the
center, sometimes even past it because of phase differences.

> This assumes of course that one doesn;t have a special design mic system
> like the Soundfield handy...
> Or did I miss the lessaon (and yes I DID read through it all!)

Want me to go over the math? :-)


Bob
--

"Things should be described as simply as possible, but no
simpler."

A. Einstein
Anonymous
April 16, 2005 1:45:41 AM

Archived from groups: rec.audio.pro (More info?)

On 4/15/05 12:12 AM, in article d3neuv02dln@enews3.newsguy.com, "Bob Cain"
<arcane@arcanemethods.com> wrote:

>>> In an MS we conceptually have three elements,
>>
>> ...
>>
>>> M = a + b*cos(theta)

> SSJVCmag wrote:
>>>
>> (SNIP THE PHOTOCOPY FROM THE THEORY BOOK...)
>
> Hey! It wasn't necessasary to open a book to write that.

Somehow I KNEW you were gonna say that...
I am not worthy...


>> My Ghod Bob... What happened to
>> "Things should be described as simply as possible, but no
>> simpler." ???
>
> That's about as simple as it can be said and still be
> reasonably accurate and predictive. I could have used
> complex vector algebra. :-)

Now just STOP THAT!
You KNOW what I mean here...!!!!


>> All that just to say that "IDEALLY one should have all-dead-on-matched mic
>> elements in a coincidental array", and so you do.
>> Whew...
>
> Right. I'm just trying to show exactly what the effects of
> them not being matched are. It is not all that well known
> that the logical angle formed by L and R is dependant on the
> matching and I'm not sure anyone has noticed the waggle
> effect that comes from phase differences.

Classical recordists have known this painfully for a generation or better,
whether they know the math or not... It's the curse of having Golden Ears in
the first place.


>> What bothers me is the trade-off between the reality of trying to mount 3
>> or 4 cardioid mics that way and the inherent distances (in centimeters!)
>> between the diaphragms what SHOULD be -zero- and COULD be closer by at least
>> an order of magnitude, and what that does to your resultant pattern
>> combinations...
....
>> Aren;t you better off with well-chosen mics that work
>> together well (as has been the case for over 1/2 century) with a
>> double-diaphragm bi-mic or good ribbon, than with the virtual bidirectional
>> element having a width measurable in inches?
>
> The double diaphragm bi is a really good way to get a
> matched situation if it is used for both and they are well
> matched at their fig 8 setting.

Ummmm most variable-pattern mics do NOT match well from bi to card to
pmni... That's the rub in choosing what works well. IIRR the TLM170 does
pretty well both ways (8-card and 8-omni),


> A ribbon and a condensor
> will have the problems with a frequency dependant angle that
> I described.
> I'll bet that if you set up an experiment with a compact
> drum set say 15 feet away at an angle of 50 degrees or so
> using a condenser M and a ribbon S with the usual HF rolloff
> that you will hear what I described.

Good grief yes, vastly different spectral and transient responses and...
Only thing I've had fun with doing this with ribbons is MS-Blumlein.

> Listening to the
> recording, the lower mids will be about where they are
> supposed to be and the cymbals will be much nearer the
> center, sometimes even past it because of phase differences.

Haven;t ever thought to try this (the idea itself seemed wrong and I didin;t
have the time to try it just to see)


>> This assumes of course that one doesn;t have a special design mic system
>> like the Soundfield handy...
>> Or did I miss the lessaon (and yes I DID read through it all!)
>
> Want me to go over the math? :-)

Don;t mind the math, just that here reading is not the situation where I
tend to spend time going over it at the speed -I- need to get even the
better part of it! I still sit down and go over the Blumlein AES preprints
every couple years to get further on in really getting the whole of it all.
Anonymous
April 16, 2005 1:45:42 AM

Archived from groups: rec.audio.pro (More info?)

SSJVCmag wrote:

>>Hey! It wasn't necessasary to open a book to write that.
>
> Somehow I KNEW you were gonna say that...
> I am not worthy...

It's funny, when someone writes down a bit of mundane
algebra/trig it somehow seems more intimidating than someone
writing about years of experience. The math part is the
easy part. At least it takes a whole lot less time to get
under your belt than gaining a working knowledge does. They
can be really complementary, though, which I am sure you
know well, and is why I spend time writing the formality here.

> Ummmm most variable-pattern mics do NOT match well from bi to card to
> pmni... That's the rub in choosing what works well. IIRR the TLM170 does
> pretty well both ways (8-card and 8-omni),

Right. The omni component and the fig 8 component of a
variable mic will have quite different responses and because
the variability is gained by varying the mix of them you
will see what you say. Inside, though, there is a matched
fig 8 in both of the M and S (if the same mic is used) which
is what will give the frequency independant reproduction
angle when using the combo regardless of the setting of the M.

>>I'll bet that if you set up an experiment with a compact
>>drum set say 15 feet away at an angle of 50 degrees or so
>>using a condenser M and a ribbon S with the usual HF rolloff
>>that you will hear what I described.
>
> Good grief yes, vastly different spectral and transient responses and...
> Only thing I've had fun with doing this with ribbons is MS-Blumlein.

We're really on the same page, then.


Bob
--

"Things should be described as simply as possible, but no
simpler."

A. Einstein
Anonymous
April 16, 2005 6:36:35 AM

Archived from groups: rec.audio.pro (More info?)

On 4/15/05 7:23 PM, in article d3pidd09ji@enews1.newsguy.com, "Bob Cain"
<arcane@arcanemethods.com> wrote:


> We're really on the same page, then.

Knew that.

Though I would like you to hit me again with digging at why I'm really
knee-jerk-against the concept of a cobbled-8 (from a pair of cards) with the
diaphragms necessarily some 2" or more apart and you don;t seem to mind it.
Anonymous
April 16, 2005 6:36:36 AM

Archived from groups: rec.audio.pro (More info?)

SSJVCmag wrote:

> Though I would like you to hit me again with digging at why I'm really
> knee-jerk-against the concept of a cobbled-8 (from a pair of cards) with the
> diaphragms necessarily some 2" or more apart and you don;t seem to mind it.

Because if stacked in a vertical line they are coincident
where it counts, in the horizontal plane. The separation
you refer to will only have an effect on sound coming from
signifigantly off that plane. This is mostly ambience and
for that, the coloration and image shifting due to lack of
coincidence is much less important IME. Given that, it is
generally less expensive to "cobble" together three good
card capsules than to rig two dual diaphragm variable
pattern capsules (and I've never seen a small diameter dual
diaphragm.) OTOH, the dual diaphragm capsules have better
low frequency response than single diaphragm cardiods. (I
think they have less proximity effect too but about that I'm
less than certain without looking up the math. :-)

Having said all of what's been said, I've come to prefer XY
(vertically stacked) to MS because two well matched cardiod
capsules are all that is necessasary for image stability and
the math is pretty much the same so that the L and R can be
decoded to M and S to be re-encoded at a different logical
angle just like real MS can. I can't tell by listening
whether a recording is MS or XY unless I can hear the image
instability that comes from mismatched MS.


Bob
--

"Things should be described as simply as possible, but no
simpler."

A. Einstein
Anonymous
April 16, 2005 6:36:36 AM

Archived from groups: rec.audio.pro (More info?)

SSJVCmag wrote:
> On 4/15/05 7:23 PM, in article d3pidd09ji@enews1.newsguy.com, "Bob Cain"
> <arcane@arcanemethods.com> wrote:
>
>
>
>>We're really on the same page, then.
>
>
> Knew that.
>
> Though I would like you to hit me again with digging at why I'm really
> knee-jerk-against the concept of a cobbled-8 (from a pair of cards) with the
> diaphragms necessarily some 2" or more apart and you don;t seem to mind it.

BTW, who are you anyway? Unless you've done a nym shift you
seem to be sorta new here (hardly new, just new here.) It
doesn't really matter, I guess, but you're prolific as well
as knowledgable and have got me really curious.


Bob
--

"Things should be described as simply as possible, but no
simpler."

A. Einstein
Anonymous
April 16, 2005 2:53:46 PM

Archived from groups: rec.audio.pro (More info?)

> the L and R can be decoded to M and S to be re-encoded at a
> different logical angle just like real MS can.

.... Interesting. Yes, that ought to work, and if you're in the digital
domain it ought to be pretty darned clean.

In fact I'm really tempted to throw together a plug-in which does this
in realtime...
Anonymous
April 16, 2005 4:43:31 PM

Archived from groups: rec.audio.pro (More info?)

Joe Kesselman wrote:
>> the L and R can be decoded to M and S to be re-encoded at a different
>> logical angle just like real MS can.
>
>
> ... Interesting. Yes, that ought to work, and if you're in the digital
> domain it ought to be pretty darned clean.
>
> In fact I'm really tempted to throw together a plug-in which does this
> in realtime...

Don't mean to discourage you, but Waves has one. I can't
remember the name offhand but it also has other really nice
functions for panning and precedence that seem to work well.
Whether the input track is MS or XY is selectable.


Bob
--

"Things should be described as simply as possible, but no
simpler."

A. Einstein
Anonymous
April 16, 2005 5:45:15 PM

Archived from groups: rec.audio.pro (More info?)

On 4/16/05 2:19 AM, in article d3qapj013jh@enews1.newsguy.com, "Bob Cain"
<arcane@arcanemethods.com> wrote:

> SSJVCmag wrote:
>
>> Though I would like you to hit me again with digging at why I'm really
>> knee-jerk-against the concept of a cobbled-8 (from a pair of cards) with the
>> diaphragms necessarily some 2" or more apart and you don;t seem to mind it.
>
> Because if stacked in a vertical line they are coincident
> where it counts, in the horizontal plane.

Ah, you;re talking like an XY head-above-head positioning but swung out to a
full 180? The bodies blocking one side don;t bother?

The separation
> you refer to will only have an effect on sound coming from
> signifigantly off that plane. This is mostly ambience and
> for that, the coloration and image shifting due to lack of
> coincidence is much less important IME. Given that, it is
> generally less expensive to "cobble" together three good
> card capsules than to rig two dual diaphragm variable
> pattern capsules (and I've never seen a small diameter dual
> diaphragm.)

AKG made a fig-8 head for the 451 series...
Didn;t Neumann make something...?


> Having said all of what's been said, I've come to prefer XY
> (vertically stacked) to MS because two well matched cardiod
> capsules are all that is necessasary for image stability and
> the math is pretty much the same so that the L and R can be
> decoded to M and S to be re-encoded at a different logical
> angle just like real MS can. I can't tell by listening
> whether a recording is MS or XY unless I can hear the image
> instability that comes from mismatched MS.

And that¹s as it should be. I guess it was that my intro to MS was as
assitant in classical work in nice halls where the first choice of an MS was
an omni. We didn;t start moving to wide-card unless forced by some aspect of
the venue. Later I discovered MS-Blum and THAt has its own special place.
Anonymous
April 16, 2005 5:45:16 PM

Archived from groups: rec.audio.pro (More info?)

SSJVCmag wrote:

> Ah, you;re talking like an XY head-above-head positioning but swung out to a
> full 180? The bodies blocking one side don;t bother?

I was thinking of side addressed mics where the bodies are
vertically oriented for MS or XY. Bodies will always give
diffraction but intuitively it seems that it will be more
benign in that configuration. I'd be cautious of using end
addressed mics for MS, but for XY the fact that each card's
body is in its null somewhat ameliorates the diffraction
effects.

> AKG made a fig-8 head for the 451 series...
> Didn;t Neumann make something...?

Could well be that I've missed it but I've not seen a dual
diaphragm variable pattern capsule with a small diaphragm.
I'd love to find some such capsules to try in a Soundfield
configuration.


Bob
--

"Things should be described as simply as possible, but no
simpler."

A. Einstein
Anonymous
April 16, 2005 5:53:40 PM

Archived from groups: rec.audio.pro (More info?)

On 4/16/05 2:26 AM, in article d3qb7r02p8j@enews2.newsguy.com, "Bob Cain"
<arcane@arcanemethods.com> wrote:

> BTW, who are you anyway?

I'M BATMAN!
No sorry...

>Unless you've done a nym shift you
> seem to be sorta new here (hardly new, just new here.) It
> doesn't really matter, I guess, but you're prolific as well
> as knowledgable and have got me really curious.

Oh dear... I've been a source of help and bother to folks here for years!
Not hardly new here but the tag changes. (I frustrate folks by being both
annoyed in the extreme by being spam-mined off NG addresses, and at the same
time really ignorant about real use of a computer and the net).
Life hit on several planes a couple years back and I NEEDED to just stop the
habit and cold-turkeyed. It's creeping back though...
Mike, Roger, Brian, Hank, several others seem find me tolerable company
despite my foibles... even in social situations!

John V
Anonymous
April 16, 2005 6:06:30 PM

Archived from groups: rec.audio.pro (More info?)

On Sat, 16 Apr 2005 12:43:31 -0700, Bob Cain
<arcane@arcanemethods.com> wrote:

>
>
>Joe Kesselman wrote:
>>> the L and R can be decoded to M and S to be re-encoded at a different
>>> logical angle just like real MS can.
>>
>>
>> ... Interesting. Yes, that ought to work, and if you're in the digital
>> domain it ought to be pretty darned clean.
>>
>> In fact I'm really tempted to throw together a plug-in which does this
>> in realtime...
>
>Don't mean to discourage you, but Waves has one. I can't
>remember the name offhand but it also has other really nice
>functions for panning and precedence that seem to work well.
> Whether the input track is MS or XY is selectable.

Yep the Waves plugin is really good, it's called the "Shuffler".

Al
Anonymous
April 17, 2005 12:32:26 AM

Archived from groups: rec.audio.pro (More info?)

playon wrote:

>>Don't mean to discourage you, but Waves has one. I can't
>>remember the name offhand but it also has other really nice
>>functions for panning and precedence that seem to work well.
>> Whether the input track is MS or XY is selectable.
>
>
> Yep the Waves plugin is really good, it's called the "Shuffler".

Right! There's also the "S1-Imager" but the "S1-Shuffler"
additionally has Blumlein shuffling included.


Bob
--

"Things should be described as simply as possible, but no
simpler."

A. Einstein
Anonymous
April 17, 2005 3:29:24 AM

Archived from groups: rec.audio.pro (More info?)

Bob Cain wrote:
> Don't mean to discourage you, but Waves has one.

Well, yes. Wave$ has lot$ of excellent $oftware, and it'$ worth what
they charge for it.

But this is a function that doesn't require their level of
sophistication, and it'd be a good excuse to learn how to code to that API.

> it also has other really nice functions for panning and
> precedence that seem to work well.

That's probably reason to use theirs rather than a homebrew.
Anonymous
April 17, 2005 7:19:40 PM

Archived from groups: rec.audio.pro (More info?)

On 4/16/05 11:29 PM, in article d3sl65$3a3$1@domitilla.aioe.org, "Joe
Kesselman" <keshlam-nospam@comcast.net> wrote:

> Bob Cain wrote:
>> Don't mean to discourage you, but Waves has one.
>
> Well, yes. Wave$ has lot$ of excellent $oftware, and it'$ worth what
> they charge for it.
>
> But this is a function that doesn't require their level of
> sophistication,

Ah...ah...ah... If I read the name right, 'the shuffler' implies that
they;re going the extra mile and not just doing the decoding but also the
dedicated adjustmets , 'shuffling' that goes with it...

>...and it'd be a good excuse to learn how to code to that API.

Never a bad exercise


>
>> it also has other really nice functions for panning and
>> precedence that seem to work well.
>
> That's probably reason to use theirs rather than a homebrew.
!