Yes. MSED [[url=http://[/URL]="http://www.voxengo…"]Audio mid-side encoder-decoder plugin (AU, VST) - MSED - Voxengo[/]="http://www.voxengo…"]Audio mid-side encoder-decoder plugin (AU, VST) - MSED - Voxengo[/]

It can reverse the encoding.

Free?

Since it is free...is it that good?

Yes. MS encoding/decoding is not exactly rocket science. I have several mastering plugs that can extract just a center channel as well. In your case however you just need a straight decode.

SO exactly what option do I have to hit on this thing. I know it is probably obvious but I do request your patience.

Click on where it says . It will give you a drop down that asks whether you want to or

Utilizing MS matrixing can be very effective in processing all kinds of stereo information. You can either create the MS matrix or decode the MS matrix. Many pieces of software such as Adobe Audition & many others include a " channel mixing" feature for stereo tracks. These usually include some kinds of presets to go to or from MS. This technique was not only used for microphone recording but also to encode FM stereo & special analog NAB, continuous loop cartridge recorders. This technique for analog recording prevented certain head azimuth alignment problems On playback. There are certain situations when you're given a stereo feed that isn't correct. Sometimes mistakes are made and you might find a soloist in your left or right channel instead of stereo center. Utilizing MS encoding or decoding can certainly straighten out a crooked and/or lopsided stereo mix.

Back slashed & white washed
Mx. Remy Ann David

OK. So I selected decode and now the left channel is much higher than the right - I am quite sure that all of the action at the venue I was at was just on the left. Any ideas?

I would not use a plugin. The sum of the left and right should give you the "center" channel. The difference (sum of the right plus the inverted left) should give you the "side" (or its inverse). You will have to mess with the phase (polarity) of the side and perhaps its volume.

Of course now that it is un-encoded you could just flip the polarity.

Care to explain

What part?

Of course now that it is un-encoded you could just flip the polarity.

I have gotten it pretty close with the plugin but the volume is really off

Well within either Stuido One or Logic or whatever DAW is a button that says phase or polarity or something like that. If you hit that the left right image will be reversed. That was your main concern. Within in many editors you can split a stereo track into it's left right components. This makes it much easier to deal with too. Then there are numerous stereo plugins that can do similar things.

Now, if you think the image was reversed then you had the VP88 pointed 180 degrees reversed. Generally the "side" microphone points to the left as the audience looks at the stage. This is also known as stage right.

One of the reasons many of us manually decode MS within a DAW is to control all the variables ourselves rather than rely on the mic or recorder matrix.

So I have been using the suggested filter and i think I got it. The only issue is that the "side" - right channel is much lower than the left. If I boost it then the correlation meter goes to hell. Any ideas?

Use your ears and not a meter.

I have been told this before and I usually do but I don't want to shoot myself in the foot and find out everything is wrong later becuase of it

Ok. Here's the deal. When you are using a mid side microphone or two near coincident microphones as mid side, you aren't going to get perfectly equal levels. How could you. In one ear you have mid and in the other you have side. One preamp knob only controls the center or mid. One preamp knob covers the side. Basically three channels but you can't adjust left to right during the recording. That has to be done after the fact just like you are doing. If the sound stage was not centered on the microphone then either left or right will HAVE to be hotter than the other side. Logic.

I understand this but there is a significant difference (I believe i mentioned this above) than what was at the location. There were variances but we are talking over 20db

Hard to say. I wasn't running your gear. Also, even if your mic is pointed center and your sound source is not center then you will have issues. I recorded a concert in January and there were six different compositions of various instruments. Each one slid further left until ALL the musicians were left of center stage. Fortunately this ain't my first rodeo and I know how to run automation. It's also key to know what your mic patterns are good for and what they aren't.

You're sweating 20dB. That's poor mic position or poor encoding/decoding. Or you have decoded to mid and side and the difference IS the difference between the two mic elements.

Not to put myself on a pedestal, but I know poor mic position is not the culprit. I believe it is poor encoding/decoding and i am just trying to figure this out.

I've been reading through the posts in this thread and scratching my head over what is actually being discussed. It seems there are two issues: (1) M-S encoding and decoding, and (2) what to do about a heavy left-side orientated recording.

(1) M-S encoding and decoding to/from L-R is just one process. The same algorithm works in both directions, and it does not need to know whether it is currently going from L-R to M-S or vice versa. This applies whether we are talking about using a block of hardware or a software plug-in.

(2) The heavy left-side recording should really be sorted out at the L-R level, although I have have success in manipulating M-S channels to achieve the bulk of the required centering before final adjustment of the L and R levels. All the required information from the sound sources is present, but the trouble comes in dealing with natural reverb and other effects that take on strange properties if pushed though a process that is there for a specific spatial correction of a sound source.

Insofar as there is any detailed advice that can be given, make sure that you have a correctly-functioning M-S encoder/decoder that would place the L and R outputs convincingly given a M-S source that represents a reasonable spread across the acoustic stage. If the source is lop-sided to the left, for example, it is possible to unbalance the M-channel feed into the L and R outputs so that the R output receives more of the M channel before adding in the phase/antiphase components of the S channel. A different S-level balance will be needed for the two L and R outputs. The overall output levels will be wrong after going through this process, but they can be adjusted conventionally.

As an observation, this correction is a method that is probably better performed in hardware, as most plug-ins will not be expecting deliberate L-R unbalance in this way.

Thank you for that. My major misunderstanding is that the recorder did not show these heavily uneven levels and it appears that the other side is largely composed of white noise and not actual sound. How do you propose that I take the mid channel and throw it into the R channel. At what point should I do this in my chain of doing things?

I'm not sure what you mean by "the other side". If you made a conventional M-S recording, one track on the recorder would have the M information on it and the other would have the S information. Are you saying that when you apply an M-S decoder to the output of the recorder, you get a respectable level in the L channel but just noise in the R channel?

What do the individual M and S channels sound like? Don't try to play them together as though they were stereo, but play each separately as a mono track into both your monitors. Evaluate the source level relative to the extraneous noise in each of the tracks, but avoid changing the replay levels between listening to the M then the S.

Was the Output Mode Selection switch set to MS on the VP88 when you made the recording?

Recently I bought a new device, the H4n and I turned on the M/S Matrix feature. I did not know it applied to the XLR inputs so I M/S Matrixed a source that was already perfect stereo.

Bos, I'm pretty sure this is exactly what happened. The VP88 was already matrixing the MS and then the Zoom matrixed it again. I'm guessing this isn't completely fixable-at least not easily by an amateur. By the time the gain was boosted on the side channel enough to sound ok at the recording source it would have really whacked out the signal levels due to the dual processing.

This is the bit I'm not clear about from the OP's post, and why I asked how the VP88 matrixing switch was set. If the VP88 was switched to produce an M-S output, it's correct to have M-S decoding enabled on the H4N. The resulting recordings would have been regular L-R stereo, and it would be then just a matter of what to do with the off-centre grouping.

If the problem is that the VP88 was actually switched to one of its L-R settings, then enabling the M-S matrix on H4N would have resulted in M-S recordings, needing M-S decoding on replay as I described previously.

I've only once used a VP88, and I found its L-R stereo output was unconvincing in any of the three width settings. I got better imaging by taking the raw M-S output from the mic and doing my own decoding.

Agreed about the VP88. I had one for a couple weeks then sold it. I'm still looking for a nice stereo condenser that doesn't sale for 12k.

To undo M/S encoding:

1. Sum both M/S channels for one of your stereo channels.

2. Invert one M/S channel and sum with the other M/S channel for the other stereo channel.

Which one is left and which one is right depends on which input is treated as "side" by the encoding. It may be necessary to invert polarity of one channel.

In order to fix the problem, a better understanding of what software you are using would be helpful. If the ZOOM was set to MS, the proper procedure would be to utilize the MS decode feature in your software. It sounds like you don't even know what you're looking at in your software? So let's get on the ball here instead of saying "Dr. it hurts when I do this".

Regular MSerer
Mx. Remy Ann David

Sorry it has taken me so long to reply but the holidays have had me busy. I did not make a convention M-S recording. I have a Left and Right source feeding a recording that thought it was getting a M-S source. The output mode was set to stereo on the VP88

Sounds like you got it.

Well as mentioned - the M-S decoding has produced quite uneven levels....to the extent that the right channel has almost no sound.

I am sorry if you find my descriptions sub par, I have been attempting to be as accurate and complete as possible. I have attempted to do this in Studio One, Soundtrack Pro and Logic. I am most familiar with the former two.

After all of this I am starting to believe that I did not record it re-M/Sed and that I somehow screwed with it to make it mono, as when I play the original without any filters that is what it sounds like and looks like on the phase meter. I think I am going to leave it like that and mix it with the other mics and hope for the best. I do thank you all for your time and I am sure all of this information will come in use in the future.

It is possible to record mono with the VP88. It is also possible the mic was too far away and so the image would seem more mono than stereo. You could try a stereo expander plugin but I think you may have to write this one off and combine it like you mention.

Interesting. Other than just taking one channel how could one record mono with the VP88? Also, would you suggest a filter for stereo expander?

huntson, post: 369614 wrote: After all of this I am starting to believe that I did not record it re-M/Sed and that I somehow screwed with it to make it mono, as when I play the original without any filters that is what it sounds like and looks like on the phase meter.

Can you describe what you saw when using the phasemeter? Was it a solid line top right to bottom left? Did you try the phasemeter on the decoded L and R signals, or were you only using the supposed M and S channels (i.e. the output of the recorder)?

I don't believe the VP88 can be set to give the same mono signal on both output channels. Even on its narrowest width stereo setting, the L and R outputs would show significant width in a phasemeter.

Neither did I but it looks mono to me. The first pic is the supposed M/S recording. The second pic is after I have applied a M/S decode filter  

You could try rotating the first one 90 degrees. You'll need a proper stereo phase shifter plugin for that.

huntson, post: 369748 wrote: Neither did I but it looks mono to me. The first pic is the supposed M/S recording. The second pic is after I have applied a M/S decode filter

A conventional phase plot of an M-S encoded channel pair would have the M channel on the horizontal (X) axis and the S channel on the vertical (Y) axis. If this is what you are showing, the implication is that you have mostly S-channel and very little (but not zero) M-channel.

In this type of M-S phase plot, a source source position that was completely L and no R would show as a line top right to bottom left, whereas a completely R sound would show as bottom right to top left. A completely mono sound would be horizontal, and a fully difference sound would be vertical (similar to what you show).

When decoded to L-R stereo and displayed on a phase plot, an M-S pair with only M present would show as a line top left to bottom right, and an M-S pair with only S present will show as a line bottom right to top left. We see neither of these in your second plot, so if nothing else, this indicates your M-S decoding is suspect.