mid/side splitter/combiner

Thanks all for the cool info. Nice techniques !

OK, so it'd be interesting to add 'to-m/s' & 'from-m/s' shells around processors.

Sorry if it's inappropriate :grin: , but speaking of stereo-width etc it makes me think of the Beh.
EX1 Edison. Various summing/subtraction going in in that box as well,
but in combination with filtering - to avoid losing bass I understand
(never seen a schematic though).

So all this encoding/decoding-stuff around processors is without any filtering, correct ?
(apart from the EQ you deliberately apply that is)

Bye,

Peter

Wow - I've been using MS for live recordings here and there, but didn't realize there was such an interesting set of techniques for mastering. :grin: :guinness:

[quote author="clintrubber"]Thanks all for the cool info. Nice techniques !
Sorry if it's inappropriate :grin: , but speaking of stereo-width etc it makes me think of the Beh.
EX1 Edison. Various summing/subtraction going in in that box as well,
but in combination with filtering - to avoid losing bass I understand
(never seen a schematic though).

So all this encoding/decoding-stuff around processors is without any filtering, correct ?
(apart from the EQ you deliberately apply that is)
[/quote]

I recently took another look at my old Edison. To me it appears as if the box does nothing but MS-process the signal i.e adjust the M and S level. The LMF-knob seems to do what in MS processing is called shuffling, i.e. boost the low mids in the side signal. Usually the low mids are reduced in proportion in the mid signal. But I don't think the Edison does that. I've read somewhere that the Edison uses allpass filters for manipulating the stereo image, but I'm not sure that's true. At least you can do very similar things by just clever MS processing + shuffling/filtering. Nonetheless the Edison is handy for its phase correlation meter, which is all I need it for these days.
BTW there is a German made and a Chinese made version. I don't know if there's a difference between the two except in cosmetics. Mine is the older German made one.

from Rossi:

I recently took another look at my old Edison. To me it appears as if the box does nothing but MS-process the signal i.e adjust the M and S level. The LMF-knob seems to do what in MS processing is called shuffling, i.e. boost the low mids in the side signal. Usually the low mids are reduced in proportion in the mid signal. But I don't think the Edison does that. I've read somewhere that the Edison uses allpass filters for manipulating the stereo image, but I'm not sure that's true. At least you can do very similar things by just clever MS processing + shuffling/filtering.

Click to expand...

I'd love to see a schematic but as we know... Can't remember having opened up this box.
But what's your conclusion, it is or isn't suited for some M/S-use ?

Nonetheless the Edison is handy for its phase correlation meter, which is all I need it for these days.

Click to expand...

I guess that phase meter feature gets the most use indeed. I have it connected to the inserts of a pair of subgroups so that I easily can route anything to it.

BTW there is a German made and a Chinese made version. I don't know if there's a difference between the two except in cosmetics. Mine is the older German made one.

Click to expand...

No idea how old mine is - I guess it's a Chinese version. It feels solid enough.

Bye,

Peter

[quote author="clintrubber"]
I'd love to see a schematic but as we know... Can't remember having opened up this box.
But what's your conclusion, it is or isn't suited for some M/S-use ?
[/quote]

I don't think we'll ever see a schematic. I looked inside mine, and there was nothing too fancy in it. Just opamps & stuff. No special chips or anything.
I think as far as stereo imagers go, it does its job. The question is if you need an imager at all. I used to use it on a subgroup that I routed my delay returns to. I usually used these "spaced" delays on synth pads and stuff like that. But you have to be careful to not produce too much off phase signal.

Since I've gone digital, I rarely use the edison anymore. But you can do similar stuff with plugins. There's a freeware imager by MDA. You can actually use two of them: one that converts LR to MS and another for MS back to LR and do some processing inbetween the two. EQ, dynamics whatever, provided the inbetween-plugins have a separate left and right channel (which then becomes mid and side).

resurrection....

thinking about doing jakob's circuit with some ll1517's I have.... will they work well in this application? I can't exactly tell from the datasheet what the z is, or if it has enough henries not to load at line level.

if they will work, is there anything special i should do as far as loading either side of the transformer is concerned?

Just some thoughts on the use of MS processing in mastering as it is my main gig these last few years. MS can be a great rescue tool for mixes you receive that have problems such as

-vocal too low or too high
-washy/phasey stereo action from acoustic guitars/keys that need controling
-de-essing vocals in mastering
-tightening up the low end in a controlled fashion
-any situation where you need access to a single part of the mix (such as I want the electrics up and more crunchy while leaving the vocal alone)
-good old stereo widening / and at least as often the opposite, making tracks more mono compatible

but....

I would never use it on my own mixes because there are so many other more effective ways of getting the job done. I could never imagine mixing through an MS chain though some folks do,

carry on,
Ruairi

I don't know if I should post my question here. I'm fairly new to this amazing forum.

I have an old coleman audio mbp4. It's a 4 channel vu meter. I would have 1 & 2 for L-R metering and I would like 3 & 4 to display M-S. I would not monitor through this vu meter.

I would like to know if I could use the OEP A262A7E for this. http://www.oep.co.uk/pdf/A262A7E.pdf

Or another fairly cheap alternative. I would like to implement it as described by Jakob with 1 + 1 : 1 + 1.

Zander

PS If I should post this elsewhere please tell me. I read the forum rules.

Zander said:

I don't know if I should post my question here. I'm fairly new to this amazing forum.

I have an old coleman audio mbp4. It's a 4 channel vu meter. I would have 1 & 2 for L-R metering and I would like 3 & 4 to display M-S. I would not monitor through this vu meter.

I would like to know if I could use the OEP A262A7E for this. http://www.oep.co.uk/pdf/A262A7E.pdf

Click to expand...

I don't really like the idea of using transformers to eavesdrop on a noble audio signal. Transformers present a non-linear load that creates a noticeable amount of distortion. Considering that unbuffered VU-meters also introduce their own sort of distortion (due to the internal rectifier), I can't recommend this.
Particularly in regard the OEP xfmr is an output transformer.
An active high-impedance L-R to M-S matrix is the way to go. Cheap and easy too.

abbey road d enfer said:

Zander said:

I don't know if I should post my question here. I'm fairly new to this amazing forum.

I have an old coleman audio mbp4. It's a 4 channel vu meter. I would have 1 & 2 for L-R metering and I would like 3 & 4 to display M-S. I would not monitor through this vu meter.

I would like to know if I could use the OEP A262A7E for this. http://www.oep.co.uk/pdf/A262A7E.pdf

Click to expand...

I don't really like the idea of using transformers to eavesdrop on a noble audio signal. Transformers present a non-linear load that creates a noticeable amount of distortion. Considering that unbuffered VU-meters also introduce their own sort of distortion (due to the internal rectifier), I can't recommend this.
Particularly in regard the OEP xfmr is an output transformer.
An active high-impedance L-R to M-S matrix is the way to go. Cheap and easy too.

Click to expand...

Could you suggest said matrix?

I think the meters are buffered. They have panels with differential input IC's for the bridging input.
Which feed to the meter buffers on the main board. Hope this helps.

I would not monitor through these audio signals...

one option:

http://www.ka-electronics.com/KAelectronics/MS_Matrix/MS_Matrix.htm

Zander said:

abbey road d enfer said:

Zander said:

I don't know if I should post my question here. I'm fairly new to this amazing forum.

I have an old coleman audio mbp4. It's a 4 channel vu meter. I would have 1 & 2 for L-R metering and I would like 3 & 4 to display M-S. I would not monitor through this vu meter.

I would like to know if I could use the OEP A262A7E for this. http://www.oep.co.uk/pdf/A262A7E.pdf

Click to expand...

I don't really like the idea of using transformers to eavesdrop on a noble audio signal. Transformers present a non-linear load that creates a noticeable amount of distortion. Considering that unbuffered VU-meters also introduce their own sort of distortion (due to the internal rectifier), I can't recommend this.
Particularly in regard the OEP xfmr is an output transformer.
An active high-impedance L-R to M-S matrix is the way to go. Cheap and easy too.

Click to expand...

Could you suggest said matrix?

I think the meters are buffered. They have panels with differential input IC's for the bridging input.
Which feed to the meter buffers on the main board. Hope this helps.

I would not monitor through these audio signals...

Click to expand...

Ok. So there is probably a place somewhere in your mixer where the L and R signals are present in unbalanced form.
You just need to connect one leg of the differential input to L and the other to R, and voila! You may have to adjust the sensitivity of the meter to suit your needs.

abbey road d enfer said:

Ok. So there is probably a place somewhere in your mixer where the L and R signals are present in unbalanced form.
You just need to connect one leg of the differential input to L and the other to R, and voila! You may have to adjust the sensitivity of the meter to suit your needs.

Click to expand...

Which leg do you mean? The opamps are ua741cn from st micro electronics. I don't follow you. I've been staring for hours at the electronics. I'm sorry.

Zander

https://www.dropbox.com/s/o2chiu359peapim/IMG_2542.jpg

Zander said:

abbey road d enfer said:

Ok. So there is probably a place somewhere in your mixer where the L and R signals are present in unbalanced form.
You just need to connect one leg of the differential input to L and the other to R, and voila! You may have to adjust the sensitivity of the meter to suit your needs.

Click to expand...

Which leg do you mean? The opamps are ua741cn from st micro electronics. I don't follow you. I've been staring for hours at the electronics. I'm sorry.

Zander

https://www.dropbox.com/s/o2chiu359peapim/IMG_2542.jpg

Click to expand...

Google for "differential amp". You will see there are two inputs: a positive V+in and a negative V-in. They are what I call legs.
The 741 (or any other opamp) can be made into a differential amp but surrounding it with 4 resistors. You cannot use it directly without these additional components.
Differential amp is the basis for "active balanced inputs".

abbey road d enfer said:

Google for "differential amp". You will see there are two inputs: a positive V+in and a negative V-in. They are what I call legs.
The 741 (or any other opamp) can be made into a differential amp but surrounding it with 4 resistors. You cannot use it directly without these additional components.
Differential amp is the basis for "active balanced inputs".

Click to expand...

I did a quick sketch of the differential amp.

v+ = non-inverting input and v- = inverting input?

Zander

Zander said:

abbey road d enfer said:

Google for "differential amp". You will see there are two inputs: a positive V+in and a negative V-in. They are what I call legs.
The 741 (or any other opamp) can be made into a differential amp but surrounding it with 4 resistors. You cannot use it directly without these additional components.
Differential amp is the basis for "active balanced inputs".

Click to expand...

I did a quick sketch of the differential amp.

v+ = non-inverting input and v- = inverting input?

Zander

Click to expand...

V+=positive supply rail, typically about 12-18V
V-= negative supply rail, same as V+ but opposite polarity.

You don't need R5, C3, C4 & C7

abbey road d enfer said:

V+=positive supply rail, typically about 12-18V
V-= negative supply rail, same as V+ but opposite polarity.

You don't need R5, C3, C4 & C7

Click to expand...

It's the design as it is now. I would like to know which to connect where to implement m/s encoding. That's why I drew this up for you.

Much appreciated!!

Zander

I've asked some help because I couldn't figure it out with the knowledge I have. Somebody made me this on a stripboard to insert between the diff amp and the vu buffer.

But there is a problem.

When I align the vu's:
L: +4dBu = 0VU = R
L + R (two sine waves at 1kHz) -> M = 0vu ; S = -inf
L - R (two sine waves at 1 kHz with polarity reversed on 1 channel) -> M = -inf ; S = 0VU

> Everything works as expected

When I try only the L channel needle around M = +/- -6VU ; S = -5VU
When I try only the R channel needle around M = +/- -6VU ; S = -10VU

>Should be a drop of -6dB?

>More concerned about the difference when only L or R signal applied.

If someone can check if this circuit is correct for the application.

Here is the schematic he gave me. It's showing only the Side/Diff circuit. (Can upload the Mid/Sum circuit if somebody is interested)

Zander said:

When I try only the L channel needle around M = +/- -6VU ; S = -5VU
When I try only the R channel needle around M = +/- -6VU ; S = -10VU

>Should be a drop of -6dB?

Click to expand...

Yes. I think the linearity of the meter buffer is at cause (or the meter's graduation).

>More concerned about the difference when only L or R signal applied.

Click to expand...

That is not normal at all.

If someone can check if this circuit is correct for the application.

Click to expand...

Nothing wrong with the circuit. Where is it connected in the system? You mention diff amps; where are they? Can you draw a diagram of your system , including the type of connections (balanced/unbalanced)

abbey road d enfer said:

Nothing wrong with the circuit. Where is it connected in the system? You mention diff amps; where are they? Can you draw a diagram of your system , including the type of connections (balanced/unbalanced)

Click to expand...

>The diff amps are connected via a card slot and are floating above the main board on the sides. The dotted green and black lines was the original connection to the main board.
>Please ask if something isn't clear.