[Gary Eickmeier]

I haven't read about this before, but I have experienced it. Wonder what
anyone knows about it.

I recorded a high school jazz band in MS, with the mikes quite close - like
2 rows of seats back and low in height, maybe ear height. Some strange
effects when I got home and started processing the MS into stereo. Some
directionality was wrong. The drum kit should be well to the right, but it
appeared in the center. An electric piano was even further right, and it
appeared just to the right of the drums. All of this happened on the first 3
tracks, then it all straightened out and the directionality was fantastic.
The drums and piano appeared well outside the speaker locaitions -
accurately! Listening in surround.

On another occasion, I was just trying to do a sound check for
directionality, so I walked around my MS pair and announced left channel,
right channel, center, but on playback it just didn't work. My theory is
that when you are so close to a microphone pair, both mikes hear it
omnidirectionally, the sound not having established a strong direction yet.
Improvement in this situation in the second half is probably because I was
tweeking the gains between jazz sets.

Gary Eickmeier

[Tom McCreadie]

"Gary Eickmeier" wrote:

I recorded a high school jazz band in MS, with the mikes quite close - like
2 rows of seats back and low in height, maybe ear height. Some strange
effects when I got home and started processing the MS into stereo. Some
directionality was wrong. The drum kit should be well to the right, but it
appeared in the center. An electric piano was even further right, and it
appeared just to the right of the drums. All of this happened on the first 3
tracks, then it all straightened out and the directionality was fantastic.
The drums and piano appeared well outside the speaker locaitions -
accurately! Listening in surround.
Sound sources from the extreme edges of the podium will often fall in the
so-called ambiphonic sectors of the mic array, from which they generate - after
MS to XY decoding, of course - opposite polarity signals at the L- and R-
speakers.

These "phasey" signals give confusing psychoacoustic signals to the human brain:
at first it may sound wide, pleasant and ethereal; sometimes it is felt as woozy
and unfocused and indeed it often can collapse into a centered image or even
shift to the other speaker side.

In short, all bets are off with imaging of sound sources located in the
ambiphonic regions, and people may even experience/interpret the signals in
diffeent ways.

On another occasion, I was just trying to do a sound check for
directionality, so I walked around my MS pair and announced left channel,
right channel, center, but on playback it just didn't work. My theory is
that when you are so close to a microphone pair, both mikes hear it
omnidirectionally, the sound not having established a strong direction yet.
Improvement in this situation in the second half is probably because I was
tweeking the gains between jazz sets.

If you increase the M gain of the MS pair, or narrow the included angle of a
coincident XY pair, you can often tighten up the imaging as such actions
decrease the angular width of the two left and right ambiphonic sectors. When
setting up, you could perhaps give more thought to the angle that the podium
edges are subtending to your mic array.

[Mike Rivers[_2_]]

On 12/12/2015 1:31 AM, Gary Eickmeier wrote:
I recorded a high school jazz band in MS, with the mikes quite close - like
2 rows of seats back and low in height, maybe ear height. Some strange
effects when I got home and started processing the MS into stereo. Some
directionality was wrong. The drum kit should be well to the right, but it
appeared in the center. An electric piano was even further right, and it
appeared just to the right of the drums. All of this happened on the first 3
tracks, then it all straightened out and the directionality was fantastic.

When you get too close with an M-S array, you can get some odd imaging
effects. However, since the problem "fixed itself" after three songs, I
suspect that there may have been a problem with the side channel - the
mic, the cable, the preamp. Listen to the two channels (mid and side,
not left and right) and see if one of them changed significantly at the
point when the problem "fixed."

--
For a good time, visit http://mikeriversaudio.wordpress.com

[Scott Dorsey]

Gary Eickmeier wrote:
I recorded a high school jazz band in MS, with the mikes quite close - like
2 rows of seats back and low in height, maybe ear height. Some strange
effects when I got home and started processing the MS into stereo. Some
directionality was wrong. The drum kit should be well to the right, but it
appeared in the center. An electric piano was even further right, and it
appeared just to the right of the drums. All of this happened on the first 3
tracks, then it all straightened out and the directionality was fantastic.
The drums and piano appeared well outside the speaker locaitions -
accurately! Listening in surround.

And did it change substantially as you adjusted the M/S ratio? If so, it
is likely due to the microphones being too widely spaced apart. If not,
it is likely due to standing wave problems in the room. Move the mikes
a few inches and it will change.

This is why having good monitoring in the field is so important.

On another occasion, I was just trying to do a sound check for
directionality, so I walked around my MS pair and announced left channel,
right channel, center, but on playback it just didn't work. My theory is
that when you are so close to a microphone pair, both mikes hear it
omnidirectionally, the sound not having established a strong direction yet.

No. That is almost certainly due to the microphones not occupying the same
place in space.

Improvement in this situation in the second half is probably because I was
tweeking the gains between jazz sets.

Did you move anything? What gains did you change?

Have you tried listening to this on standard speakers?
--scott


--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Gary Eickmeier]

Hi Scott -

"Scott Dorsey" wrote in message
...

And did it change substantially as you adjusted the M/S ratio? If so, it
is likely due to the microphones being too widely spaced apart. If not,
it is likely due to standing wave problems in the room. Move the mikes
a few inches and it will change.

I think all MS pairs are coincident. I suspend mine one above the other. And
no, I couldn't seem to fix it by adjusting the ratio. And yes, as I can best
recall, I did move the mikes a few inches back after the rest of the
audience came in and sat right in front of me.

This is why having good monitoring in the field is so important.

Here are some of the problems of amateur field recording (!). The band is
positioned tight up against the first row of seats, near the right side of
the auditorium. Don't ask why. The audience, mostly parents, will actually
sit there in the first row available to be close to their kids and of course
to annoy me. I place my mike stand in between my knees and hold the base
down with my foot so it doesn't fall forward with the slanted floor and
gravity. The mikes are right in the middle of the audience, with
pop-clappers all around me, each one trying to out clap the others whenever
a kid finishes a solo. I keep the mikes at ear height so that they don't get
in the eye line of the people behind me. I am there just for the experiment
of it, and I like the jazz band's music. They were terrific, and that is
about my only opportunity to record them. Now, if I give the director a copy
of the recording, I might be able to get them to let me do a more
professional recording during rehearsals. I would love that.


On another occasion, I was just trying to do a sound check for
directionality, so I walked around my MS pair and announced left channel,
right channel, center, but on playback it just didn't work. My theory is
that when you are so close to a microphone pair, both mikes hear it
omnidirectionally, the sound not having established a strong direction
yet.

No. That is almost certainly due to the microphones not occupying the
same
place in space.

Again, MS pairs are coincident.


Improvement in this situation in the second half is probably because I was
tweeking the gains between jazz sets.

Did you move anything? What gains did you change?

I try to record with equal gains in MS, because I can adjust the ratios
later. Therefore, I am getting a little more of the S channel than normal,
which would give MORE separation, rather than less. But of course I adjust
as required in post.


Have you tried listening to this on standard speakers?
--scott

Yes, on the editing bench. Three small computer speakers. Same problem. Out
in the theater room I can listen in surround or standard stereo, but the
surround shows the directional problems faster than the stereo. It was a
most delightful effect to hear the drum kit and piano further around to the
side on my right, and the applause all around me, rather than collapsing
everything to the front speakers or the front wall.

I have been doing MS for a while now, recording the signals raw in the field
and then manually mixing down to stereo in post. I think I will do it again
tomorrow with the concert band and the closer small group pre show, which
will be a semicircle of clarinets. Will let you know if all problems go away
with sensible distances!

Thanks,
Gary Eickmeier

[Mike Rivers[_2_]]

On 12/12/2015 1:31 AM, Gary Eickmeier wrote:
Some strange
effects when I got home and started processing the MS into stereo. Some
directionality was wrong. The drum kit should be well to the right, but it
appeared in the center. An electric piano was even further right, and it
appeared just to the right of the drums. All of this happened on the first 3
tracks, then it all straightened out and the directionality was fantastic.

Did I misunderstand this? Was the problem in the recording, and
something changed after the first three songs without you doing anything?


--
For a good time, visit http://mikeriversaudio.wordpress.com

[John Williamson]

On 12/12/2015 22:13, Mike Rivers wrote:
On 12/12/2015 1:31 AM, Gary Eickmeier wrote:
Some strange
effects when I got home and started processing the MS into stereo. Some
directionality was wrong. The drum kit should be well to the right,
but it
appeared in the center. An electric piano was even further right, and it
appeared just to the right of the drums. All of this happened on the
first 3
tracks, then it all straightened out and the directionality was
fantastic.

Did I misunderstand this? Was the problem in the recording, and
something changed after the first three songs without you doing anything?


If the S mic is one with switchable patterns, maybe it wasn't in full
figure 8 mode until it dried out in the heat of the hall, or a switch
suddenly made contact at a critical temperature or moisture level ?

--
Tciao for Now!

John.

[Scott Dorsey]

Gary Eickmeier wrote:
Hi Scott -

"Scott Dorsey" wrote in message
...

And did it change substantially as you adjusted the M/S ratio? If so, it
is likely due to the microphones being too widely spaced apart. If not,
it is likely due to standing wave problems in the room. Move the mikes
a few inches and it will change.

I think all MS pairs are coincident. I suspend mine one above the other. And
no, I couldn't seem to fix it by adjusting the ratio. And yes, as I can best
recall, I did move the mikes a few inches back after the rest of the
audience came in and sat right in front of me.

How coincident is concident? A quarter inch? A sixteenth of an inch?
Two inches?

This is why having good monitoring in the field is so important.

Here are some of the problems of amateur field recording (!). The band is
positioned tight up against the first row of seats, near the right side of
the auditorium. Don't ask why. The audience, mostly parents, will actually
sit there in the first row available to be close to their kids and of course
to annoy me. I place my mike stand in between my knees and hold the base
down with my foot so it doesn't fall forward with the slanted floor and
gravity.

Take over a dressing room or a bathroom backstage, your life will be a
thousand times easier. Get a proper mike stand and rope off the area
around the stand. Sandbag it.

If you have reflection problems in the room like you describe, sometimes
getting the microphone up higher can help that a lot. Not always, though..
and you have to listen and see.

No. That is almost certainly due to the microphones not occupying the
same
place in space.

Again, MS pairs are coincident.

Unfortunately this is not physically possible. There is only "mostly
coincident" which can be pretty good with some designs but not really
very good with others.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[PStamler]

Have you tried listening to this on standard speakers?
--scott

Yes, on the editing bench. Three small computer speakers. Same problem.


*Three* speakers? How were they placed?

Peace,
Paul

[Gary Eickmeier]

"PStamler" wrote in message
...

Have you tried listening to this on standard speakers?
--scott

Yes, on the editing bench. Three small computer speakers. Same problem.


*Three* speakers? How were they placed?

Left, center, and right. Sitting on the tabletop of the editing bench. Why,
what did you expect? The result, anyway, was the same in editing and in the
theater room.

Gary

[Gary Eickmeier]

As far as I can determine, no switch changes or electrical problems. As
Scott has suggested, it was probably acoustical, which would be a lot harder
to trace down. I will just shrug my shoulders on this until next time. My S
mike is on figure 8 with the positive direction left, and my M mike is on
cardioid, pointing forward. At home, the signals all looked good and strong,
no cutouts or weak gains. But the damn imaging changed after the first 3
numbers.

What I do at home is put the M on the multitrack timeline twice, then put
the S signal on once, and the inverted S signal on once. I now have four
tracks to mix down to stereo. The Left channel is composed of the M and the
S signals, both panned left. The Right channel is made from the M and the
inverted S signal, both panned Right.

One of the beauties of MS is that you usually get perfect channel balance
because each channel is made by splitting M and S evenly. Changing the ratio
between M and S just changes the stereo spread, not the channel balance. If,
during recording, I feel a need to change the gain on one of the mikes to
even things up, all that changes in the result is the ratio of M and S,
relatively innocuous.

Gary


----- Original Message -----
From: "John Williamson"
Newsgroups: rec.audio.pro
Sent: Saturday, December 12, 2015 5:25 PM
Subject: MS Proximity Problem


On 12/12/2015 22:13, Mike Rivers wrote:
On 12/12/2015 1:31 AM, Gary Eickmeier wrote:
Some strange
effects when I got home and started processing the MS into stereo. Some
directionality was wrong. The drum kit should be well to the right,
but it
appeared in the center. An electric piano was even further right, and it
appeared just to the right of the drums. All of this happened on the
first 3
tracks, then it all straightened out and the directionality was
fantastic.

Did I misunderstand this? Was the problem in the recording, and
something changed after the first three songs without you doing anything?


If the S mic is one with switchable patterns, maybe it wasn't in full
figure 8 mode until it dried out in the heat of the hall, or a switch
suddenly made contact at a critical temperature or moisture level ?

--
Tciao for Now!

John.

[[email protected]]

Did you change the gain during recording?

[geoff]

On 13/12/2015 6:22 PM, Gary Eickmeier wrote:
"PStamler" wrote in message
...

Have you tried listening to this on standard speakers?
--scott

Yes, on the editing bench. Three small computer speakers. Same problem.


*Three* speakers? How were they placed?

Left, center, and right. Sitting on the tabletop of the editing bench. Why,
what did you expect? The result, anyway, was the same in editing and in the
theater room.

Gary




Three speaker though ???

M&S stereo needs only two. What is the third doing ?

geoff

[Mike Rivers[_2_]]

On 12/13/2015 12:24 AM, Gary Eickmeier wrote:

As far as I can determine, no switch changes or electrical problems. As
Scott has suggested, it was probably acoustical, which would be a lot harder
to trace down.

Can you think of anything that might have happened? You'd probably
notice if the drummer or piano moved, and surely the walls didn't move,
but how about audience? From your description, you were sitting pretty
much in the thick of it. Did someone sitting near your mic that might
have blocked some sound get up and move?

What I do at home is put the M on the multitrack timeline twice, then put
the S signal on once, and the inverted S signal on once. I now have four
tracks to mix down to stereo. The Left channel is composed of the M and the
S signals, both panned left. The Right channel is made from the M and the
inverted S signal, both panned Right.

That can certainly work, and it's what I did with a real mixer and on a
computer before I had a program that lets me treat a two-channel file as
an M-S pair and translates it to stereo. In fact, my TASCAM DR-44WL will
do that with its internal mixer, giving me a control to adjust the
stereo width.




--
For a good time, visit http://mikeriversaudio.wordpress.com

[Scott Dorsey]

Gary Eickmeier wrote:
"PStamler" wrote in message
...

Have you tried listening to this on standard speakers?
--scott

Yes, on the editing bench. Three small computer speakers. Same problem.


*Three* speakers? How were they placed?

Left, center, and right. Sitting on the tabletop of the editing bench. Why,
what did you expect? The result, anyway, was the same in editing and in the
theater room.

Wait, wait... now you're processing something again to create a false
center channel. What if you just play it back on standard two channel
systems?
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Gary Eickmeier]

Well I...... I don't think the center channel is doing anything fancy to the
signal such as center steering in DPL - but I have a center speaker in the
theater too, so I think it is harmless, or gives a better idea of the mix
than only two speakers, which would shift imaging as I move my head. There
is no gain control for the center as opposed to the L and R.

I don't worry too much about the mixing speakers - they just give me a clue
that everything is working, then I look at the meters for channel balance
and loudness, and take the result out to the theater to evaluate everything
else. Then back into the mix to correct as required.

Gary



"Scott Dorsey" wrote in message
...
Gary Eickmeier wrote:
"PStamler" wrote in message
...

Have you tried listening to this on standard speakers?
--scott

Yes, on the editing bench. Three small computer speakers. Same problem.


*Three* speakers? How were they placed?

Left, center, and right. Sitting on the tabletop of the editing bench.
Why,
what did you expect? The result, anyway, was the same in editing and in
the
theater room.

Wait, wait... now you're processing something again to create a false
center channel. What if you just play it back on standard two channel
systems?
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Gary Eickmeier]

I don't think my Zoom H6 can do the MS mixing with external mikes, but of
course it has the MS microphone accessory with variable ratio if you like
the internal mike; Makober, yes I did change the gains a little during and
between recordings, but I don't think that would account for moving the drum
set from extreme right to center.

I am wondering if, by using the MS technique, I am riding right on the edges
of the patterns of the fig 8 mike and the cardioid center mike for a source
like where that drum kit was? In that case, maybe just a slight change in
the direction the bracket was pointing would shift the image from the fig 8
more to the center.

In any case, here I go again to the concert hall for today's session with
the big concert band. Big fun!

Gary


----- Original Message -----
From: "Mike Rivers"
Newsgroups: rec.audio.pro
Sent: Sunday, December 13, 2015 8:19 AM
Subject: MS Proximity Problem


On 12/13/2015 12:24 AM, Gary Eickmeier wrote:

As far as I can determine, no switch changes or electrical problems. As
Scott has suggested, it was probably acoustical, which would be a lot
harder
to trace down.

Can you think of anything that might have happened? You'd probably notice
if the drummer or piano moved, and surely the walls didn't move, but how
about audience? From your description, you were sitting pretty much in the
thick of it. Did someone sitting near your mic that might have blocked
some sound get up and move?

What I do at home is put the M on the multitrack timeline twice, then put
the S signal on once, and the inverted S signal on once. I now have four
tracks to mix down to stereo. The Left channel is composed of the M and
the
S signals, both panned left. The Right channel is made from the M and the
inverted S signal, both panned Right.

That can certainly work, and it's what I did with a real mixer and on a
computer before I had a program that lets me treat a two-channel file as
an M-S pair and translates it to stereo. In fact, my TASCAM DR-44WL will
do that with its internal mixer, giving me a control to adjust the stereo
width.




--
For a good time, visit http://mikeriversaudio.wordpress.com

[Mike Rivers[_2_]]

On 12/13/2015 11:41 AM, Gary Eickmeier wrote:

I am wondering if, by using the MS technique, I am riding right on the
edges of the patterns of the fig 8 mike and the cardioid center mike for
a source like where that drum kit was? In that case, maybe just a slight
change in the direction the bracket was pointing would shift the image
from the fig 8 more to the center.

You should be able to cover about 150 degrees, though your mic should be
centered on the band. Also, understand that M-S doesn't have as accurate
imaging as X-Y and it doesn't take but a small shift side to side or
front to back to through things out of whack. That's why Scott keeps
telling you that you need good monitoring.

Since you're splitting the side mic to two tracks in your DAW program,
you might try changing the level of one of those tracks to see if you
can get your recording to sound more like it looked on stage.


--
For a good time, visit http://mikeriversaudio.wordpress.com

[Mike Rivers[_2_]]

On 12/13/2015 11:47 AM, Gary Eickmeier wrote:
I don't worry too much about the mixing speakers - they just give me a clue
that everything is working, then I look at the meters for channel balance
and loudness, and take the result out to the theater to evaluate everything
else. Then back into the mix to correct as required.

If you depend on the meters for channel balance using an M-S setup with
one channel being the mid mic and the other being the side mic, you
aren't really able to tell anything except that it's working and that
you're not overloading.

I had a peek at the Zoom H6 manual, and it looks like stereo decoding is
pretty much restricted to using their M-S mic. It appears that you have
some fixed options for S level, but those may not appear if you don't
have the M-S mic plugged in. You might consider cobbling up an outboard
M-S decoder with a headphone amplifier so you can hear stereo when using
your own mics in M-S configuration. Wes Dooley (AEA, the ribbon mic
folks) used to make one. Wonder if he still have any on hand. Jensen
Transformers had an application note for building one out of a couple of
dual winding transformers. Neither is a cheap solution, but unless you
can monitor in real left/right stereo, you're just guessing if your mics
are in the right place - and that goes for X-Y stereo, too.


--
For a good time, visit http://mikeriversaudio.wordpress.com

[Frank Stearns]

Mike Rivers writes:

On 12/13/2015 11:47 AM, Gary Eickmeier wrote:
I don't worry too much about the mixing speakers - they just give me a clue
that everything is working, then I look at the meters for channel balance
and loudness, and take the result out to the theater to evaluate everything
else. Then back into the mix to correct as required.

If you depend on the meters for channel balance using an M-S setup with
one channel being the mid mic and the other being the side mic, you
aren't really able to tell anything except that it's working and that
you're not overloading.

I had a peek at the Zoom H6 manual, and it looks like stereo decoding is
pretty much restricted to using their M-S mic. It appears that you have
some fixed options for S level, but those may not appear if you don't
have the M-S mic plugged in. You might consider cobbling up an outboard
M-S decoder with a headphone amplifier so you can hear stereo when using
your own mics in M-S configuration. Wes Dooley (AEA, the ribbon mic
folks) used to make one. Wonder if he still have any on hand. Jensen
Transformers had an application note for building one out of a couple of
dual winding transformers. Neither is a cheap solution, but unless you
can monitor in real left/right stereo, you're just guessing if your mics
are in the right place - and that goes for X-Y stereo, too.

That is an expensive HW solution. What's the "software" solution missing? (M plus
S plus S dupliciated, polarity on the dupe flipped).

I've also found the soft solution very easy to calibrate. Mute M, pan S+ and S- to
mono; you should get complete cancellation (if not, check EQ and levels on the S+
and S- channels carefully). Then, you group the S channels to keep in them in
lock-step, pan them back out hard L and R, and you're ready to go. Adjust the M and
S relative levels to get the desired image.

To address the original question about the changing image... doesn't take more than
a dB or two of mis-cal to screw things up... I wonder if the cal had somehow been
out. Could have even been something simple like a level difference between the two
segments. Maybe if Gary re-cal'd for both segments...

Frank
Mobile Audio
--

[jason]

On Sun, 13 Dec 2015 14:27:51 -0500 "Mike Rivers" wrote
in article
If you depend on the meters for channel balance using an M-S setup with
one channel being the mid mic and the other being the side mic, you
aren't really able to tell anything except that it's working and that
you're not overloading.


If XY can be converted to MS, is there a compelling reason to produce the
original recording in MS given the mic issues?

[Gary Eickmeier]

Jason, just the flexibility of the post processing of MS. If you record XY,
you must decide on the included angle, which maybe wouldn't always be 90
degrees - would it?

Frank, what are you meaning by calibration? The gain settings? No, my gain
knobs are not accurate, compared to each other. I set gains by the meters,
not the knob numbers. This may create a slight problem in MS recording
because all I know to do is make the gains equal during recording, which may
not be accurate as far as the stereo spread when all is said and done. My
starting point in the mix is to set the M and S gains to be equal, then to
move the ratios up or down to get center fill right.

And Jason, I like MS because it is a poor man's really good surround sound
encoder, if you listen to playback in DPL-II surround. I can definitely hear
sounds that were off to the sides or even to the rear coming back to me
accurately on playback. Startling sometimes.

This last session, of the big band Christmas concert, went through like
butter. I just set up my MS mikes, set levels to what I know will work, and
kept them there the whole concert. I use a second recorder to capture the
house mikes to mix the singers' voices in directly, and both recorders
stayed in sync with each other for over 2 1/2 hours.

Haven't gotten to play it back yet, but the disc is about 78 minutes and
should have plenty of surround sound in it.

Gary



"Jason" wrote in message
...
On Sun, 13 Dec 2015 14:27:51 -0500 "Mike Rivers" wrote
in article
If you depend on the meters for channel balance using an M-S setup with
one channel being the mid mic and the other being the side mic, you
aren't really able to tell anything except that it's working and that
you're not overloading.


If XY can be converted to MS, is there a compelling reason to produce the
original recording in MS given the mic issues?

[Peter Larsen[_3_]]

On 14-12-2015 08:44, Gary Eickmeier topposted and confused the thread,
albeit in a easily fixed manner:

"Jason" wrote in message

If XY can be converted to MS, is there a compelling reason to
produce the original recording in MS given the mic issues?

No. The dimension that matters is btw. not measured in inches, but in
wavelengths and his mics, as I recall them, are quite large objects when
trying to emulate points. It would not be my first choice, but then we
all do things differently.

Jason, just the flexibility of the post processing of MS. If you record XY,
you must decide on the included angle, which maybe wouldn't always be 90
degrees - would it?

No. It should be whatever works and gives a stable center image. If you
want the benefit of MS processing you convert to MS in your Audition2
with the channel mixer and do whatever special magic you need to do or
just alter the levels and then use the inverse preset to convert back.
It should be transparent, but do it on a 32 bit file as you are likely
to "do things" while in MS.

I think I did post the preset values once, let me know if I didn't and
I'll look them up, the adobiots didn't think them necessary in newer
versions.

Note however that your strategy of doing in it multitrack mode generally
is a good idea because all post operations can be done in one combined
mathematical operation. One of the big mastering guys recommended aiming
for that when passing through here some years ago.

Generally I find that the S channel should be between a couple of dB
lower in level than the M channel and equal to it for a sane stereo
image. If that doesn't work, then there is something else wrong with the
recording, in which case I have occasionally lowered the S channel to
about -4 dB relative to the M channel.

MS recording is btw. easy to monitor sans conversion option, position
your mics for a good M signal with proper front to rear balance of the
ensemble and proper ensemble to room balance.

From your recording scenarios it appears that you need a potion or
charm that gives you the power to move chairs. Or to screw hooks into
the wall or ceiling and arrange for safe mic suspension.

Gary

Kind regards

Peter Larsen

[Tom McCreadie]

Mike Rivers wrote:

Also, understand that M-S doesn't have as accurate
imaging as X-Y
Really, Mike, how on earth can you justify that? An M-S would in general
actually give a more accurate representation of centered images. As ever, of
course, the imaging quality depends on the degree of off-axis
response-raggedness of real life mics.

Since you're splitting the side mic to two tracks in your DAW program,
you might try changing the level of one of those tracks to see if you
can get your recording to sound more like it looked on stage.
Gary has this tendency for going for convoluted workflows :-) ...for a start,
generating X-Y from M-S would be a lot simpler in post by just running the raw
M-S signal through the free DAW plug-in: "Voxengo MSED".

[Mike Rivers[_2_]]

On 12/14/2015 6:23 AM, Tom McCreadie wrote:
Mike Rivers wrote:

Also, understand that M-S doesn't have as accurate
imaging as X-Y
Really, Mike, how on earth can you justify that? An M-S would in general
actually give a more accurate representation of centered images. As ever, of
course, the imaging quality depends on the degree of off-axis
response-raggedness of real life mics.

And that's exactly my point. A mid mic that has a little bump in its
polar response on one side of center that isn't matched on the other
side, or a side mic that isn't perfectly symmetrical (a number of modern
ones are intentionally built that way) will throw off the image
accuracy. It takes a lot of care in placement, which means careful
listening, to get it right. On the other hand, reasonably well matched
cardioid mics aren't hard to find, and for the kind of recordings that
Gary is making, he's likely to get more consistent results.



--
For a good time, visit http://mikeriversaudio.wordpress.com

[Mike Rivers[_2_]]

On 12/13/2015 6:06 PM, Frank Stearns wrote:
Mike Rivers writes:

I had a peek at the Zoom H6 manual, and it looks like stereo decoding is
pretty much restricted to using their M-S mic.
.. . . . . .

You might consider cobbling up an outboard
M-S decoder with a headphone amplifier so you can hear stereo when using
your own mics in M-S configuration.

That is an expensive HW solution. What's the "software" solution missing?

Something to run the software on. Gary is recording with a hardware
recorder, not a computer. He has the wrong recorder for this job, unless
there's been an update to the Zoom H6 that allows stereo monitoring of
an M-S pair connected to the external mic inputs. I can do that on my
TASCAM DR-44 (and I think DR-40 as well - I don't remember).

To address the original question about the changing image... doesn't take more than
a dB or two of mis-cal to screw things up... I wonder if the cal had somehow been
out.

He could have bumped a knob, or his mic stand might have tilted a bit.
It didn't sound like he had a very secure setup. But when he's
recording, he has only two controls, gain for the mid mic and gain for
the side mic. If it was just a matter of a gain change, he could
compensate for that when mixing the tracks to stereo. There's no
"calibration" involved in capture if you're not going to monitor in stereo.



--
For a good time, visit http://mikeriversaudio.wordpress.com

[Mike Rivers[_2_]]

On 12/14/2015 2:44 AM, Gary Eickmeier wrote:
Jason, just the flexibility of the post processing of MS. If you record XY,
you must decide on the included angle, which maybe wouldn't always be 90
degrees - would it?

Actually, if you're using cardioid mics, you really do always want them
to be at 90 degrees. You move closer or further away from the source so
that it all "fits" in the stereo directivity pattern. You can get too
close (hole in the middle) or too far (too much ambiance relative to the
source level). That's why you have to listen to what you're recording.

However, when it comes to processing, M-S and X-Y are completely
reversible. You can convert an X-Y recording to its equivalent mid and
side components, process those as if you started with an M-S mic setup,
and then put it back to stereo.

Frank, what are you meaning by calibration? The gain settings? No, my gain
knobs are not accurate, compared to each other. I set gains by the meters,
not the knob numbers. This may create a slight problem in MS recording
because all I know to do is make the gains equal during recording, which may
not be accurate as far as the stereo spread when all is said and done.

How you set the mid and side levels _when recording_ has no bearing on
the stereo width. That's determined by how you mix the mid and +/- side
mics. Do you not understand this?

What Frank means by "calibration" is having the "in phase" and "out of
phase" side signals equal in amplitude. You're doing this by using the
polarity invert function in your software. The two side signals, unless
you fool with them, will be at the same level. If you want to adjust the
level of the side signal, you need to adjust both signals equally.
Otherwise, your left/right balance will be wrong.

The usual way to do this in a DAW program is to "group" the two side
signals in the software mixer. Pan one full left, the other to full
right, and control them both with a single software fader. Then start
your "mix to stereo" with just the mid channel and listen on speakers as
you bring up the level of the side signals. Stop when you get the width
you want. If you go too far, you'll hear a hole in the middle. If you
prefer, you can do this the other way - start listening to the two side
signals in the left and right speakers. It will sound like it's trying
to tear your face down the middle. Then bring up the mid mic level until
the center fills in.

If the stereo level gets too hot, then reduce the mid and side levels so
that you keep the mix from clipping.



--
For a good time, visit http://mikeriversaudio.wordpress.com

[Tom McCreadie]

Peter Larsen wrote:


Generally I find that the S channel should be between a couple of dB
lower in level than the M channel and equal to it for a sane stereo
image. If that doesn't work, then there is something else wrong with the
recording, in which case I have occasionally lowered the S channel to
about -4 dB relative to the M channel.

Yes. and one can get a more quantitative handle on this by first mapping the MS
over to its theoretically equivalent "virtual XY pair" then using the calculator
on the Sengpiel website to estimate the approximate SRA's of such an XY pair.

As illustration, I include some scenarios below. It's clear that the SRA of the
virtual XY array increases as the S gain decreases. Gary would avoid
complications if he ensured that the Orchestra Angle (how the players are spread
across the podium) doesn't spill beyond the SRA of the mic array.
=========
S mic gain relative to cardioid M gain:
1) -6dB
= virtual X-Y pair with:
polar pattern, V = 0.41 + 0.59.cos.theta
mic type: narrow cardioid, at XY mic included angle = 90°
= estimated Stereophonic Recording Angle (SRA) = 154°

2) -3dB
= virtual X-Y pair with:
polar pattern, V = 0.37 + 0.63.cos.theta
supercardioid; mic included angle = 110°
= SRA 115° *

3) 0dB
= virtual X-Y pair with:
polar pattern, V = 0.31 + 0.69.cos.theta
supercardioid; mic included angle 127°
= SRA 98° **

4) +3dB
= virtual X-Y pair with:
polar pattern, V = 0.25 + 0.75.cos.theta
hypercardioid; mic included angle 141°
= SRA 61°

5) +6dB
= virtual X-Y pair with:
polar pattern, V = 0.20 + 0.80.cos.theta
hypercardioid; mic included angle 152°
= SRA 46°


N.B * 115° value based on pattnrn V = 0.344 + 0.656.cos.theta
** 98° value based on pattnrn V = 0.41 + 0.59.cos.theta
============

[Frank Stearns]

"Gary Eickmeier" writes:

Jason, just the flexibility of the post processing of MS. If you record XY,
you must decide on the included angle, which maybe wouldn't always be 90
degrees - would it?

Frank, what are you meaning by calibration? The gain settings? No, my gain
knobs are not accurate, compared to each other. I set gains by the meters,
not the knob numbers. This may create a slight problem in MS recording
because all I know to do is make the gains equal during recording, which may
not be accurate as far as the stereo spread when all is said and done. My
starting point in the mix is to set the M and S gains to be equal, then to
move the ratios up or down to get center fill right.

My assumption has been that you're recording raw M and raw S, and *not* decoding to
L and R at the time of recording. Rather, the "conversion" to L & R is done in post
using:

- a mid channel, panned center (likely a cardioide but I have used omnis to good
effect, particularly those with some HF directionality)

- a side channel, initially recorded with a Fig 8 mic, panned hard left,

- an exact copy of the side channel but with the polarity flipped, panned hard
right. That copy might be a polarity flip on duplicated data, polarity flip on a
channel strip Y'd from the raw S channel, and so on -- whatever method you use to
get a second duplicate channel off the source S, just flipped.

It's important to "calibrate" the S+ and S- channels. When temporarily panned
center, you should have complete cancellation and hear nothing. If you hear
something faintly coming through, the S+ and S- channels are out of cal. Adjust
levels and EQ to get the best possible cancellation. This is much more likely a
problem when decoding with hardware (channel strips of a console) because of minute
variances in component tolerances among channels.

Also, it's nice to have field monitoring while still maintaining that raw M & S
recording, so you've perhaps set up channel strips in your field monitor console as
noted above.

As far as the relative level of M and S, it does not take much S to get a nice
stereo spread so often the S faders will be 6-15 dB lower than M.

As I've probably commented before, I at one time used M-S extensively but like X-Y
and to a lesser extent ORTF found that M-S does not offer the depth image possible
with 50 cm-spaced omnis.

I still use M-S when overdubbing group vocals in multiple passes. I can record each
pass with a nice stereo image, but then in post will often throw away the S channels
and pan the M of each pass differently to get a more striking L/R spread of those
vocals.

YMMV.

Frank
Mobile Audio

--

[[email protected]]

To the OP,

was there ANY kind of AGC engaged while you were recording?

Mark

[Peter Larsen[_3_]]

On 14-12-2015 13:52, Mike Rivers wrote:

On 12/13/2015 6:06 PM, Frank Stearns wrote:

I had a peek at the Zoom H6 manual, and it looks like stereo decoding is
pretty much restricted to using their M-S mic.

You might consider cobbling up an outboard
M-S decoder with a headphone amplifier so you can hear stereo when using
your own mics in M-S configuration.

That is an expensive HW solution. What's the "software" solution missing?

Something to run the software on. Gary is recording with a hardware
recorder, not a computer. He has the wrong recorder for this job, unless
there's been an update to the Zoom H6 that allows stereo monitoring of
an M-S pair connected to the external mic inputs. I can do that on my
TASCAM DR-44 (and I think DR-40 as well - I don't remember).

It doesn't matter, I made some M-S experiments with a friend and we
found it easy to get the mic position right by aiming for a good M.

To address the original question about the changing image... doesn't
take more than
a dB or two of mis-cal to screw things up... I wonder if the cal had
somehow been
out.

He could have bumped a knob, or his mic stand might have tilted a bit.
It didn't sound like he had a very secure setup. But when he's
recording, he has only two controls, gain for the mid mic and gain for
the side mic. If it was just a matter of a gain change, he could
compensate for that when mixing the tracks to stereo. There's no
"calibration" involved in capture if you're not going to monitor in stereo.

Gary has some fairly large mics, they are not adept at emulating a point
but have the advantage of selectable directionality.

Kind regards

Peter Larsen

[Peter Larsen[_3_]]

On 14-12-2015 14:45, Frank Stearns wrote:

As I've probably commented before, I at one time used M-S extensively but like X-Y
and to a lesser extent ORTF found that M-S does not offer the depth image possible
with 50 cm-spaced omnis.

I used that a lot for a chamber music festival with extremely good
results, including good tolerance of being close to the sound sources.

YMMV.

With all setups there is a learning phase, allow for it.

Frank
Mobile Audio

Kind regards

Peter Larsen

[Frank Stearns]

Peter Larsen writes:

On 14-12-2015 14:45, Frank Stearns wrote:

As I've probably commented before, I at one time used M-S extensively but like X-Y
and to a lesser extent ORTF found that M-S does not offer the depth image possible
with 50 cm-spaced omnis.

I used that a lot for a chamber music festival with extremely good
results, including good tolerance of being close to the sound sources.

Ah, yes. You've hit on one of my key reasons for liking this technique besides the
imaging. One can get close -- for lots of good detail and a fair amount of direct
signal unmolested by room problems -- yet you still have a lot of room "space" in
there as well.

Given the inherent limits of Recording As We Know It (and with even the finest
gear), this is the best of both worlds.

With all setups there is a learning phase, allow for it.

Indeed.

Frank
Mobile Audio
--

[Gary Eickmeier]

Frank - you said something that bothers me - see below - you said that

"My assumption has been that you're recording raw M and raw S, and *not*
decoding to
L and R at the time of recording. Rather, the "conversion" to L & R is done
in post
using:

" - a mid channel, panned center (likely a cardioide but I have used omnis
to good
effect, particularly those with some HF directionality)"

No, you don't pan the M channel to center, you pan the M and the straight S
to the left for the left channel, and the M and the inverted S to the right
channel.

Is there some other way to do it that you are talking about?

Gary

"Frank Stearns" wrote in message
...
"Gary Eickmeier" writes:

Jason, just the flexibility of the post processing of MS. If you record
XY,
you must decide on the included angle, which maybe wouldn't always be 90
degrees - would it?

Frank, what are you meaning by calibration? The gain settings? No, my gain
knobs are not accurate, compared to each other. I set gains by the meters,
not the knob numbers. This may create a slight problem in MS recording
because all I know to do is make the gains equal during recording, which
may
not be accurate as far as the stereo spread when all is said and done. My
starting point in the mix is to set the M and S gains to be equal, then to
move the ratios up or down to get center fill right.

My assumption has been that you're recording raw M and raw S, and *not*
decoding to
L and R at the time of recording. Rather, the "conversion" to L & R is
done in post
using:

- a mid channel, panned center (likely a cardioide but I have used omnis
to good
effect, particularly those with some HF directionality)

- a side channel, initially recorded with a Fig 8 mic, panned hard left,

- an exact copy of the side channel but with the polarity flipped, panned
hard
right. That copy might be a polarity flip on duplicated data, polarity
flip on a
channel strip Y'd from the raw S channel, and so on -- whatever method you
use to
get a second duplicate channel off the source S, just flipped.

It's important to "calibrate" the S+ and S- channels. When temporarily
panned
center, you should have complete cancellation and hear nothing. If you
hear
something faintly coming through, the S+ and S- channels are out of cal.
Adjust
levels and EQ to get the best possible cancellation. This is much more
likely a
problem when decoding with hardware (channel strips of a console) because
of minute
variances in component tolerances among channels.

Also, it's nice to have field monitoring while still maintaining that raw
M & S
recording, so you've perhaps set up channel strips in your field monitor
console as
noted above.

As far as the relative level of M and S, it does not take much S to get a
nice
stereo spread so often the S faders will be 6-15 dB lower than M.

As I've probably commented before, I at one time used M-S extensively but
like X-Y
and to a lesser extent ORTF found that M-S does not offer the depth image
possible
with 50 cm-spaced omnis.

I still use M-S when overdubbing group vocals in multiple passes. I can
record each
pass with a nice stereo image, but then in post will often throw away the
S channels
and pan the M of each pass differently to get a more striking L/R spread
of those
vocals.

YMMV.

Frank
Mobile Audio

--
.

[Frank Stearns]

"Gary Eickmeier" writes:

Frank - you said something that bothers me - see below - you said that

"My assumption has been that you're recording raw M and raw S, and *not*
decoding to
L and R at the time of recording. Rather, the "conversion" to L & R is done
in post
using:

" - a mid channel, panned center (likely a cardioide but I have used omnis
to good
effect, particularly those with some HF directionality)"

No, you don't pan the M channel to center, you pan the M and the straight S
to the left for the left channel, and the M and the inverted S to the right
channel.

Is there some other way to do it that you are talking about?

Not sure what you're asking, but electrically, a single channel panned center should
be *identical* to that same single-source channel duplicated and applied to a hard
left/hard right pan on two channels.

What do you think is happening internally at a pan pot??? When centered, it's
dumping equal signal level to the left and right summing busses (allowing for any
non-linearities in the pan pot itself. In a DAW, center is center and should be a
non-issue in terms of subtle hardware errors. But with HW, that's why you use
meters).

Panning M center does work with M-S as described; I have 100s of recordings to prove
it.

Frank
Mobile Audio
--

[Tom McCreadie]

Tom McCreadie wrote:
snip
N.B * 115° value based on pattnrn V = 0.344 + 0.656.cos.theta
** 98° value based on pattnrn V = 0.41 + 0.59.cos.theta

Oops. typo. For the nerds who read that far, the above should have been:
* 115° value based on a supercardioid pattern V = 0.344 + 0.656.cos.theta
** 98° value also based on a supercardioid pattern V = 0.344 + 0.656.cos.theta

[Mike Rivers[_2_]]

On 12/14/2015 2:07 PM, Peter Larsen wrote:
I made some M-S experiments with a friend and we found it easy to get
the mic position right by aiming for a good M.

That might work for a symphony orchestra, a solo piano, or a group
that's well balanced within itself and symmetrical. But with Gary's
amateur bands, and not monitoring when he's setting up, it's just going
to be a guess.

--
For a good time, visit http://mikeriversaudio.wordpress.com

[Scott Dorsey]

Peter Larsen wrote:
On 14-12-2015 14:45, Frank Stearns wrote:

As I've probably commented before, I at one time used M-S extensively but like X-Y
and to a lesser extent ORTF found that M-S does not offer the depth image possible
with 50 cm-spaced omnis.

I used that a lot for a chamber music festival with extremely good
results, including good tolerance of being close to the sound sources.

My objection is that the "sense of depth" that you get in the recording
is unrealistic and not like the sound in the actual hall. Lots of people
like it, though, and it's less extreme than the Mercury triad.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[geoff]

On 15/12/2015 2:26 a.m., Mike Rivers wrote:


The usual way to do this in a DAW program is to "group" the two side
signals in the software mixer. Pan one full left, the other to full
right, and control them both with a single software fader. Then start
your "mix to stereo" with just the mid channel and listen on speakers
as you bring up the level of the side signals. Stop when you get the
width you want.

Slightly O the specific T, but as an aside with MS, I find the control
is so powerful, I can never actually decide when the width is how I want
it !

geoff

[geoff]

On 15/12/2015 2:45 a.m., Frank Stearns wrote:
"Gary Eickmeier" writes:

Jason, just the flexibility of the post processing of MS. If you record XY,
you must decide on the included angle, which maybe wouldn't always be 90
degrees - would it?
Frank, what are you meaning by calibration? The gain settings? No, my gain
knobs are not accurate, compared to each other. I set gains by the meters,
not the knob numbers. This may create a slight problem in MS recording
because all I know to do is make the gains equal during recording, which may
not be accurate as far as the stereo spread when all is said and done. My
starting point in the mix is to set the M and S gains to be equal, then to
move the ratios up or down to get center fill right.
My assumption has been that you're recording raw M and raw S, and *not* decoding to
L and R at the time of recording. Rather, the "conversion" to L & R is done in post
using:

- a mid channel, panned center (likely a cardioide but I have used omnis to good
effect, particularly those with some HF directionality)

- a side channel, initially recorded with a Fig 8 mic, panned hard left,

- an exact copy of the side channel but with the polarity flipped, panned hard
right. That copy might be a polarity flip on duplicated data, polarity flip on a
channel strip Y'd from the raw S channel, and so on -- whatever method you use to
get a second duplicate channel off the source S, just flipped.

Gary, are you actually getting all muddled with this Centre Speaker thing ?

Unless it is for a 5.1 mix, or something like that, ditch the Center
speaker. MS is for stereo positioning, and a 'middle and side' speakers
are not a factor !

Even if the end result will have a Centre channel, ignore it in the
context of a stereo positioning and width achieved mixing with the MS
recording.

geoff

geoff