Hi All,

I'm exited to be part of this esteemed group! I wish I'd known about this site 2 years ago as all of you would have stopped me from making amateur mistakes...

I'm interested in M+S recording so that room ambiance can be manipulated in post production. I have been test recording with a Sony ECM-680s shotgun M+S stereo microphone attached to a Zoom H4N, to get a feel of how to go about it.

The results of my short tests are not what I expected and have raised more questions than answers.

When I started my investigation, I assumed that the following work flow was in process: the Sony ECM-680s microphone is set to stereo mode, which results in the Mid recorded on the left channel and the Side recorded on the right channel. I record an highly ambient sound source to a test file on the H4N. I open up my audio editor and import this test file, and split the 2 channels into single mono tracks. I duplicate the Right (side) channel, pan the 1st copy to the left, pan the second copy to the right and play with the volume balance of the 3 tracks to get a stereo sound I like. This process seems simple and elegant to me.

However, during my recording test , I had set the Zoom H4N to Menu => input=>MS Matrix=>On/Off to “on”, unsure of what this switch does. But Since I was M+S recording I decided to set it to “on”. Watching the level meters during the test the L (mid ) channel was at a higher volume level than the R ( side) channel as I would expect as my sound source was directly in front. I was listening via monitoring headphones and liked the ambiance I was getting.

Later on I searched exhaustively on the WWW to find out what the H4N MS Matrix on/off switch actually does. The only thing I found was from the Zoom website:

H4n Handy Recorder (http://www.zoom.co.jp/english/products/h4n/)

Quote “The built-in M/S stereo decoder allows the use of an external Mid-Side microphone setups. The H4n converts the mid and side channel signals into a conventional stereo (L/R) signal for even more versatility, and it's only found on the H4n.”

I found no other information anywhere else.

What exactly does Zoom mean? If it means that the H4n is recreating the stereo image from a M+S mic source then it is defeating the point of recording in M+S mode.

This is the 1st unanswered question I have.

As a further test I plugged in a Sony ECM-CG50 mono shotgun mic with L+ R inputs into the H4N recorder to test the difference between switching MS Matrix on & off. The first test I had it set to Off. The L & R levels were matched, which I expected. I was hearing a balanced mono sound in the headphones. The second test I turned the MS Matrix switch to On and the R (side) channel volume dropped down considerably. Now the right side was considerably quieter in the headphones.

My next test was to play back each test to see if the test files were recorded the same way as I heard them. They were. So that ruled out the MS Matrix is a Headphone/line out only switch.

BTW for all of these tests, the the H4N MS Matrix channel & level settings where all at the default values.

I cant figure out what this MS Matrix processor is doing or why it is a feature on the H4N, or how it is to be used. This is the 2nd unanswered question I have.

The H4n MS Matrix switch seems to be a useless feature to me.

As a further investigation I read thru the Sony Specification sheet for the Sony ECM-680s mic directly from:

Sony Product Detail Page ECM680S (http://pro.sony.com/bbsc/ssr/cat-audio/cat-wiredmics/product-ECM680S/)

Quote

“The MS microphone signal of the ECM-680S is internally decoded and it provides L and R (stereo) outputs.”

Is this a technical writers mistake? What exactly is meant when Sony states “internally decoded”?


It makes me wonder if I am getting a M+S image from this mic.

This is the 3rd unanswered question I have.

Anyone have any answers?

Thanks

Ray

Most important is the last point. Your M/S mic is putting out a signal that has been decoded to stereo. This is not uncommon for Sony's approach to MS - I have a different Sony mic that does the same, in either "wide" or "very wide" settings per a switch on the mic.

Although you and I think the same way, "Let's get us some MS and adjust the stereo width in post", some manufacturers think that would be too complicated for us to do. There are other benefits to MS, chief among them a seamless collapse of the stereo decode to very high-fidelity mono, no comb-filtering or other phasing effects.

Were you able to record in MS, the formula to decode is as follows:
L = M+S
R = M-S

In a modern NLE, that means doubling the S track. Pan one of them hard left. Invert the other and pan it hard right. Now, keep both these tracks at the same gain setting, and control stereo width by mixing in the M track (panned center).

I recently did some math to re-encode M/S from a recorded stereo signal (originated on MS, of course!).
If I did it right:
M = L+R
S = L-R

Didn't experiment much with the results, but they seemed promising.

About the H4n, I don't know if the MS decode matrix is applied to the incoming signal or just the monitor circuit. Both are common applications for MS recording. But, you don't want either with your mic...

Thanks Seth,

it looks like you are correct and the sony mic is doing the processing for me.

I'd hoped otherwise.

I wish to record like this Web clip demonstrated here

Created by Camtasia Studio 4 (http://www.recording-microphones.co.uk/Mid%20Side%20FLV/Mid%20Side2.html)

as that would add a powerful dimension to my recordings.

That clip is a good overview of the decode process. The same procedure is possible in almost any DAW or NLE. Note at 1:40, "switch the phase...", this is what I referred to as "invert". Properly, we both should have said "invert the phase...".

I am a big fan of stereo recording techniques, having done a little with M/S and lots with ORTF.

Most inexpensive M/S mikes do include a decode matrix such that their output is stereo. This greatly simplifies monitoring, of course, since few mixers & recorders have a decode matrix for monitoring.

However, I do have a project recorded to stereo from M/S, and believe that recode to M/S is possible. See again the recode formula above, I was put on to this by another forum member a couple years ago.

M = L+R
S = L-R

Once I've rendered the M and S tracks, I expect to be able to decode this material with stereo width control. If this works, you can do the same from your Sony mic.

To lay this out in a DAW/NLE, first separate L + R so you can pan them separately, and pan both to center. Render/output this, call it Mid.

Then, on the same track layout, invert the phase of the R track, render this, call it Side.

Now, go to the decode instructions above or to the nice little screen cam, and set up an M/S decode project and see how it works!

The math to convert M/S to L/R is basically the same as the math to convert L/R to M/S.

M+S=L M-S=R
L+R=M L-R=S

So if you take the stereo output of your mic, and turn on the "matrix" in the Zoom, then the channels recorded on the Zoom should actually be M and S.

You can confirm this with a test... preferably outdoors, or in a very large dead room, to minimize confusion from reflections and reverberation.

Talk 90 degrees to the left of the mic, then go the same distance 90 degrees to the right, and talk again. Then go the same distance directly to the front of the mic and talk again.

The first two tests should produce recorded signals that are roughly the same level. The first test will produce two tracks in phase, the second test will produce two tracks out of phase.

The third test should produce good signal on the first track of the recording (the M channel) but nearly nothing on the second track of the recording (the S channel).

Do what you want in post.