Is there a difference between Mono and Mid?

[ObsoleteAcc99]

[DELETED]

[murnau]

well there you go, you already go it. And you can solo each band to hear exactly what you are working with, and i don't know why one would consider basslane superior for this task to mix control when mix control quite frankly poos on it as it has 3 bands doing it (and expanding as well).

because i didn't know that ferez21 use mixcontrol (he mention just after my post) but anyway.. ..for the task he asked tiny basslane is just perfect + free: i personally dont need 3 bands for splitting / narrowing a certain low bassfreq.

[bbaggins]

Mid=Left+Right
Side=Left-Right

So mid has all the information from the sound so yes discarding the side leaves you with a mono signal

Are you sure Mid have all the information? i think it shouldnt have.

Remember the reason Mid/Side encoding was invented in the first place: to allow consumers to play stereophonic recordings on their monophonic record players.

When you do that, you are hearing the Mid component only, which is embedded in the lateral variations of the record groove. Your mono turntable would ignore the vertical excursions that contained the Side information and you'd only hear the Mid component, which was a genuine mono version of the record you were listening to.

[herodotus]

A great deal of confusion results from the fact that 'mid/side' properly applies not to just any stereo signal, but specifically to a stereo recording using the mid/side stereo micing technique.



The advantage over other stereo micing patterns (e.g. XY, ORTF) is that it is 100% mono compatible.

Most other uses of the term refer to a signal processor mimicking the effects of this sort of stereo field and the control it gives over the stereo image.

[Andrew Souter]

Mid=Left+Right
Side=Left-Right

So mid has all the information from the sound so yes discarding the side leaves you with a mono signal

Are you sure Mid have all the information? i think it shouldnt have.

Converting two channels to one channel can simply be done in three ways:

1) Throw away one input channel. Keep the other.
2) Evenly mix both channels into one with both channels in phase (mid)
3) Evenly mix both channels into one with one channel phase inverted (side)

(you could also unevenly mix, and then throw away one channel)

All methods potentially "loose information" depending on what was in the stereo signal to begin with.

If for example your stereo input consisted of a signal only in the right, and you keep right, you have not lost anything. If you keep left, you have lost everything.

If your stereo signal was actually two identical signals (effectively mono) and you mix L and R together to get mid, you loose nothing. (you may have some increased gain depending on if your mix rule was equal gain or equal power) If you phase invert one channel and mix (effectively the same as subtracting) you have lost everything.

If your input signal was exactly 180degrees out of phase (R = -L) and you mix both channels, you have lost everything. If you phase invert one channel and mix (effectively the same as subtracting) you have lost nothing b/c you have inverted a negative signal making it the same phase as the other channel.

More complexly, if you mix a signal that has a sineWaveA at 100hz, in the left channel, sineWaveB at 125hz in both channels, and sineWaveC at 150hz in both channels but phase inverted in the R channel, any way you covert this complex two channel signal to a one channel signal you are going to loose something... If you throw away the left channel, you have lost 100hz. If you mix both channels to get the Mid, you will loose 150Hz. If you mix to Side via phase inverting one channel and summing you will loose 125hz.

In a complex recording the phase interactions will be much more complex, and stereo signals that have an average intra-channel correlation of zero (like a reverb tail, or ambient recording) will interact in much more complicated ways when summing. You will get some kind of extremely complex comb filtering when mixing. The spectrum will change any way you do it.

And yes, you will loose something. It is unavoidable. And once lost it can not be recreated further downstream.

Whether or not this is objectionable, is debatable. Like most everything else.

[Tricky-Loops]

Here is a great article about Mid-Side-Recording:

http://www.uaudio.com/blog/mid-side-mic-recording

If you convert a stereo recording to Mid, it would be pure mono. But if you record your own stuff with a Mid mic, there would be some sonic influence from the Sides, too, (and possible phase cancellations) or can it be completely eliminated?

[gavriloP]

The standard process to convert stereo to mono (as it happens in hardware, hifi and software) is sum left and right. This iss the same thing as mid signal. This was said before and it is just like that. But of course when you are converting from stereo to mono you can also use L or R channel only for example. I do this sometimes when I want to turn stereo verb to mono or stereo piano to mono, for example. It might sound better for that occasion.

Here is another thread about this and something I wrote in it: http://www.kvraudio.com/forum/viewtopic ... 51#5094651

[djscorb]

Use MSED to encode your stereo files into two mono files; ie the Sum (L+R) and Difference (L-R).

Now, with two mono tracks in your daw you have the ability to mix some or all of the side channel together with the mono channel to preserve more of that "side" information.

This will add the decorrelated stereo material that might otherwise have got lost if you are using the SUM (or Mid) exclusively.

Mid side is indeed a microphone technique and some argue that the term mid/side for audio processing would be better named as Sum and Difference processing.

Take DMG Equilibrium for instance, you can apply sum and difference processing between user specified surround channels, so the term mid and side is not always very helpful or indeed accurate!

Cheers

Scorb

[djscorb]

Oops, I meant to add the link for MSED

http://www.voxengo.com/product/msed/

Cheers

Scorb

[Krakatau]

Mid=Left+Right
Side=Left-Right

So mid has all the information from the sound so yes discarding the side leaves you with a mono signal

you need to avoid the usual arithmetic conception or this affirmation shall be irrevocably misunderstood, mind you...

[edited] Oops, ...i did't notice Galbanum's post that knows certainly the topic in all details, my bad !



I don't know the accurate translation of the matter i've learned as graduate called "Théorie des ensembles" in "modern maths" , but referring to this, this assertion would then be correct ..!

[Keith99]

Mid=Left+Right
Side=Left-Right

So mid has all the information from the sound so yes discarding the side leaves you with a mono signal

Are you sure Mid have all the information? i think it shouldnt have.

Converting two channels to one channel can simply be done in three ways:

1) Throw away one input channel. Keep the other.
2) Evenly mix both channels into one with both channels in phase (mid)
3) Evenly mix both channels into one with one channel phase inverted (side)

(you could also unevenly mix, and then throw away one channel)

All methods potentially "loose information" depending on what was in the stereo signal to begin with.

If for example your stereo input consisted of a signal only in the right, and you keep right, you have not lost anything. If you keep left, you have lost everything.

If your stereo signal was actually two identical signals (effectively mono) and you mix L and R together to get mid, you loose nothing. (you may have some increased gain depending on if your mix rule was equal gain or equal power) If you phase invert one channel and mix (effectively the same as subtracting) you have lost everything.

If your input signal was exactly 180degrees out of phase (R = -L) and you mix both channels, you have lost everything. If you phase invert one channel and mix (effectively the same as subtracting) you have lost nothing b/c you have inverted a negative signal making it the same phase as the other channel.

More complexly, if you mix a signal that has a sineWaveA at 100hz, in the left channel, sineWaveB at 125hz in both channels, and sineWaveC at 150hz in both channels but phase inverted in the R channel, any way you covert this complex two channel signal to a one channel signal you are going to loose something... If you throw away the left channel, you have lost 100hz. If you mix both channels to get the Mid, you will loose 150Hz. If you mix to Side via phase inverting one channel and summing you will loose 125hz.

In a complex recording the phase interactions will be much more complex, and stereo signals that have an average intra-channel correlation of zero (like a reverb tail, or ambient recording) will interact in much more complicated ways when summing. You will get some kind of extremely complex comb filtering when mixing. The spectrum will change any way you do it.

And yes, you will loose something. It is unavoidable. And once lost it can not be recreated further downstream.

Whether or not this is objectionable, is debatable. Like most everything else.

Makes sense to me!

[Krakatau]

Makes sense to me!

+1

Undoubtely...

[foosnark]

I'll second the recommendation to just use BassLane. It does what you want to do in one simple, free plugin.