Working in Ozone4, toying around with the Mid/Side in the EQ and I believe in the compressor/imager as well.

Can someone recommend me some good reads about the benefits of Mid/Side mastering?

Thanks,
Nathan

M/S EQ is usually used correctively. If what you want to EQ occurs only in the center or to the far sides / out of polarity, try applying the filter only to that component. The benefit may be effective treatment of a specific problem with less impact on the mix. The downside is that it may impact the mix more, since you're changing relative EQ/levels, which may take away from the perceived mix 'integrity'. It really is different for each case.

Sweetening or creative EQing in M/S is reasonably rare here. Some common applications:
If a track could do with some sparkle / width and/or if the mix is a bit dry, you can try a high shelve boost on the S only for air / dimension / boosting-the-in-betweens. Be carful though, it's easy to overdo this.
If you have a heavy rock mix, boosting the very low mids on the S can give double tracked guitars some more balls / weight. To be used very gently and be aware that this may make it harder to achieve loudness without obvious artifacts.
If you want to give a track a bit more bite/grip, but a high mid boost makes the vocals sound too mean/nasal, try that boost on the side signal only.

Etc...

As always, what you should do depends on what you think needs doing and will be different for each new track.

Unlinked Mid/side compression is very rarely used here.

The idea is basically that impact in the mid signal will not pull down the sides, i.e. with most material, that means the sound stage gets wider with impact in the music. Sounds like a great idea... but somehow is hardly ever applicable. I do that maybe once a year.

On the corrective side of things, unlinked M/S compression is sometimes useful, usually when a far out in panorama / far out of phase component of the mix needs to be controlled separately. Especially in combination with a side chain filter, this can be useful (controlling high mid or high frequency harshness in the sides, for example). This comes up maybe once every couple of months. De-essing (i.e. narrow band compression) on the Mid only is pretty common.

PS: To answer your actual question: I think there's some stuff about unlinked M/S processing in Bob Katz's book "Mastering Audio" (well worth a read). But to be honest, M/S processing is a straightforward enough concept to be grasped via learning by doing: Turn the knobs and listen to how the sound changes. When A/B comparing before and after (even with slight EQ), always compensate change in loudness via the plugin output gain in order to make your judgement on EQ, not loudness.

M/S EQ is used pretty rarely here, too. I don't even have an analog matrix, although I'm going to build one.

Usually it's really just notching out something in either channel, usually M. A typical problem could be something poking up in the midrange, but reducing it in L/R wipes it out too much. The Sonoris LP EQ works well for those things.

From reading this, I get the impression that the Manley varimu mid side mod is not very popular?

That is an expensive unit - 4500$ for the regular version, so people buying it should be using mid side a lot, no?

Indeed it isn't that popular. But that may be partly because the text on the Manley web states that it's either the HPF mod or the M/S mod, not both, and most people need the HPF.
Nothing wrong with having an M/S option though. Had I know that it is possible to have both in one box for a 500 bucks more, I would probably have ordered it too when I was having my mods done.

Maybe David Glasser of Airshow Mastering can chime in on how / how often he tends to use the Vari-MU switched to M/S.

I use M/S EQ every day. Like it's going out of fashion.

Me too.....

Me likes the approach of M/S
like the reverbs on side and dedicated EQ

I use M/S for very specific cuts and boosts to balance a track out. Usually no more than a db, usually less, with fairly narrow Q. I use it very often for this application.

Never M/S compression... ever. Just does weird things to the balance.

Sometimes a desser on the center channel to clean up a spitty vocal.

M/S is wonderful when used responsibly and carefully.

It is indeed a vital tool for any Mastering Engineer... Sometimes it is the perfect thing to gain both width and body. I agree that it must be used in moderation so as not to throw the balance out too much as it can be also disastrous!

I often use it to add width to a fairly dry mix - especially with vocal freqs. You can really fine tune it and make a vocal come forward - more than you can with a standard EQ, since vocals are usually panned in the mix. Also to get control over wayward stereo bottom end freqs...

I have recently started running my analog chain through an M/S matrix and am generally pleased with the results. It creates a nice level of flexibility. When not needed, I simply take out the M/S matrix and I am back to L/R.

Mhh.. isn't M/S a microphone technique? The metaphor "mid/side" makes no sense for most stereo material.

When encoding a stereo signal via sum and difference (that's the correct term!), the sum does not contain the stereo center. It contains the center and the sides! Contrary to what most would expect, a fully panned event will actually appear in the sum. It's not the "mid" and effectively breaks the whole idea of mid and side. The difference doesn't really contain the side information: It describes the width, not the side!

The most funny thing around this hype is that whole companies are based on this (intentional?) misunderstanding. Using the term M/S for stereo difference encoding is a shame by itself.

Don't oversee the effects of phase distortion between both the sum and difference, messing with this relation will practically blur/destroy the original stereo positioning. E.g. by Eqing the difference information, you effectively time-shift the width information away from the original event. That is, the width of a lead synth could influence the stereo positioning of the following (mono) snare! It sounds surprising to the untrained, but its unnatural causality can be fatiguing (it simply doesn't make sense for our brain).

Linear phase EQ?

Oh yes, that helps keeping things more or less where they were!

But still, the consequences of LR difference EQing (as well as all kinds of of difference signal manipulations) are difficult to predict, even with linear phase filtering. With Linear Phase filters, the "smear" still happens, but on both sides (future and past). The difference signal is not an audio signal, it is actually extremely sensitive to the step response behaviour of the filter (i.e. the time domain becomes very important, because it directly affects stereo positioning).

Not saying you shouldn't use such techniques, just keep in mind it isn't perfect and has nothing to do with the stupid mid/side metaphor.

With an analog matrix and processing, there is no phase error other than what's created by the EQ itself.



DC

Of course, I mean the EQ's phase distortion. The sum/difference matrix by itself is lossless, no doubt.

I have one up for grabs.It is in classifieds if anybody is searching for one.

https://gearspace.com/board/gearslut...-s-master.html

Otgh, an analog matrix may not be 100% perfect, yet the digital one almost certainly is. Whether there is a audible difference is another story.


DC

I use the Fabfilter EQ and it has a Mid/Side option as well. It's pretty cool when you cut frequencies, you can keep the lows in the centre while having the higher frequencies also on the sides, on a single instrument. I guess it's good to make space

For me it just hardly ever seems to work; the result tends to sound phasey to me, even when using a linear phase EQ. I used it maybe three or four times last year. I've never tried it OTB, so maybe there is a difference there. If so I'd be interested to hear the technical reasons why.

Keep in mind, when boosting the side signal you are essentially boosting the anti-phase. The side signal disappears in mono.