naturerecordists
[Top] [All Lists]

Re: MS stereo imaging problems with MKH30/50 combo - Sounds are swap

Subject: Re: MS stereo imaging problems with MKH30/50 combo - Sounds are swap
From: "ajustend" ajustend
Date: Sat Oct 20, 2012 5:29 pm ((PDT))

John,

Yes, I also find it very strange that in all of my research of MS stereo mi=
c'ing techniques almost nothing is mentioned about stereo-image side swappi=
ng when using a super-cardioid mid. In fact, before trying it myself, the o=
nly thing I read about using a super-cardioid pattern for recording in MS w=
as that many people liked it for it's improved "focus" and off axis attenua=
tion (with the MKH50/30 combo being second in popularity to the MKH40/30).=


Being a recordist who does a ton of mono recordings for SFX purposes I thou=
ght I saw the purported value in this statement. But, it seems, the big mis=
take in my thinking (and I am curious about other's thinking on this topic =
as well) was that the Mid mic would continue to act like a forward facing m=
ic in a stereo MS setup, when it actually does NOT. When the Mid and Side s=
ignals are decoded into stereo the end result is IDENTICAL to an equivalent=
 XY setup (shown very clearly with diagrams in this paper: http://www.ribbo=
nmics.com/pdf/technique.pdf) which means that the solo Mid super-cardioid m=
ic's off axis rejection is actually only helpful when listened to alone or =
when the MS stereo signal (decoded or not) is played back in mono.

It seems to me that in stereo recording you are basically nullifying the ad=
vantage of a super-cardioid's off axis rejection and are instead only givin=
g yourself: 1) smaller recording angles (as John stated in his example of a=
 rubber band stretching), and 2) the potential for throwing rear lobe recor=
ded signals into the opposite stereo speaker during playback. Therefore, I =
really do not see any advantage of using a super-cardioid mic over a cardio=
id for coincident stereo recording.

I have been reading over the Manfred Hibbing "XY & MS Stereo Recording Tech=
niques" paper today and have been very impressed by the information it pres=
ents about these interchangeable coincident stereo setups. I have found som=
e very lucid quotes in regards to my original post:

(these are from the polar plot comparison section of the paper, looking at =
XY recording angles)

about Cardioid
"Due to the fact that cardioid microphones record sound from all directions=
 in phase the audio recorded from the rear is not side inverted."

about Super-cardioid
"Supercardioid microphones invert the phase of any signal originating 120 d=
egrees or more off the microphone axis."  Therefore, "At [XY recording] ang=
les below 120 degrees the signal recorded from the rear are side inverted a=
nd above 120 degrees they are not."

Very cool! So, as has already been mentioned in this thread (thank you Mich=
ael Raphael), it is the phase-inverted rear lobe of the super-cardioid micr=
ophone that is causing these stereo imaging problems. If I had a pair of MK=
H50's setup in an XY configuration and attempted an equivalent 360 degree w=
alk around test I would (it seems) experience the exact same thing.

-Justin


--- In  "rock_scallop" <> w=
rote:
>
> Hi Justin,
> It is interesting, in the many times super-cardiod has been mentioned on =
this group as a possible option in an MS setup, the expected convolution of=
 the stereo image has not really been explored until now.
>
> In your localization test, I heard something similar to what Jos=E9 descr=
ibed - though for the sake of interpreting the effect I will assume a theor=
etical L/R symmetry for this array configuration.
>
> The frontal stereo image becomes expanded, and the maximum perceived widt=
h occurs somewhere between 10 and 2 o'clock. At 11 to 1 o'clock the image s=
eems stretched from 60 degrees on location to around 90 degrees in playback=
. In this frontal wedge, the relative side to side localization of subjects=
 might be fairly accurate. Though I think it is like stretching a rubber ba=
nd: as the width is expanded the depth becomes thinned.   To the sides and =
back, including 10 and 2 o'clock and beyond the image becomes severely twis=
ted.
>
> All arrays distort the image in one way or another, so I am not discounti=
ng super-cardiod MS as useful technique for stereo nature recording - thoug=
h be extra careful with any content outside the angle of focus.
>
>
> John Hartog
> rockscallop.org
>
>
>
>
> --- In  "freitojos" <josefreitas81@> wro=
te:
> >
> >
> > In my stereo speakers in a equilateral triangle 2mts  90=BA speakers. 1=
2 was 0=BA in front, 1 was 45=BA my right  2 35=BA right, 3 20=BA right 4 1=
0=BA right 5 10=BA left 6 was as 2 but far away 7 was 20=BA left 8 was 25=
=BA left 9 30=BA left 10 45=BA left and 11 12 0=BA in front.
> > I hope this helps,
> > Jos=E9
> >
> > --- In  "ajustend" <ajustend@> wrote:
> > >
> > >
> > >
> > >
> > > Here is a link to the recording I made:
> > >
> > > http://soundcloud.com/justinmullens/ms-test-mkh30-50-handsofclock
> > >
> > > -Justin
> > >
> > > --- In  "J. Mullens" <ajustend@> wro=
te:
> > > >
> > > > Hi, this is my first post to Nature Recordists.
> > > >
> > > > I have just added an MKH 30 to my collection in order to expand int=
o MS
> > > > recording and am experiencing very bad imaging (side swapping) with=
 the
> > > > setup when the sound source is located in certain places behind the
> > > > microphone.
> > > >
> > > > To clarify, I am *not *trying to record a surround signal using a s=
tereo
> > > > mic setup, but I *am *interested in maintaining a correct left-righ=
t stereo
> > > > image. When I use an XY setup, or an XY mic like a BP4025, anything=
 on the
> > > > left side of the mic (in front of it or behind) will playback out o=
f the
> > > > left speaker. Using the MKH30/50 combo this is not the case.
> > > >
> > > > Here is the test I conducted: Set the microphones up in the middle =
of an
> > > > open field so that I can walk in a full circle around it. I start d=
irectly
> > > > in front of the M/S setup, about 15-20 feet away, and walk in a clo=
ckwise
> > > > direction around mic calling out each position on the clock face. 1=
2 is
> > > > directly in front, 1, is to the right of that, 3 is directly to the=
 right,
> > > > 6 is directly behind, 9 is directly to the left side, etc.
> > > >
> > > > Here's the problem, when I get to the 4 & 5 o'clock postions (rear =
right
> > > > side) the sound source swap sides and sounds like it is coming from=
 front
> > > > left. And the exact opposite happens to the 7 & 8 o'clock positions
> > > > sounding like they are coming from the front right.
> > > >
> > > > Again, any source originating in the 180 degree arc in front of the=
 setup
> > > > performs beautifully and the stereo image is perfect. It only happe=
ns to
> > > > sounds coming from behind the mics, and if a sound source is making=
 a
> > > > constant tone you can actually hear it jump diagonally across from =
one side
> > > > of the stereo field to the other as it enters these rear right & le=
ft areas.
> > > >
> > > >
> > > > After asking around it has come to my attention that the closer you=
r Mid
> > > > mic is to a pressure-gradient mic (figure 8 polar pattern), the mor=
e of a
> > > > rear lobe you have, and thus the more chance you have of that rear =
lobe
> > > > picking up sound in the opposite channel. This definitely explains =
how, and
> > > > why, I could be experiencing side swapping.
> > > >
> > > > Since I've never used an MS setup with a pure cardioid mic I am won=
dering
> > > > if the same thing happens...  For those using a Sennheiser 30/40 MS=
 setup:
> > > > does this happen? Do you have side swapping in the 4-5 o'clock and =
7-8
> > > > o'clock positions?
> > > >
> > > > Nature recording is, for me, one of the areas where a really accura=
te
> > > > stereo image in a 360 degree circle around my recording location is
> > > > critical. I am always pointing my mics at the source I am recording=
, but
> > > > having sounds behind the mic appearing in opposite channel is just =
weird.
> > > > How do the MS recordists on this forum deal with this?
> > > >
> > > > Thanks!
> > > >
> > > > -Justin
> > > >
> > > >
> > > >
> > >
> >
>







<Prev in Thread] Current Thread [Next in Thread>
Admin

The University of NSW School of Computer and Engineering takes no responsibility for the contents of this archive. It is purely a compilation of material sent by many people to the naturerecordists mailing list. It has not been checked for accuracy nor its content verified in any way. If you wish to get material removed from the archive or have other queries about the archive e-mail Andrew Taylor at this address: andrewt@cse.unsw.EDU.AU