If you want to master in M-S, you should either export your mix as L-R and use the M-S encoder at the input of the mastering stage or else export as M-S (as you have been doing) and skip the encoding at the mastering input.

What will not work is applying encoding/decoding on only one component or on both components separately. Both M-S and L-R encoded signals are vectors and must be treated as such.

Don't forget that you must always monitor in L-R as that's the way our ears are arranged.

Why on earth would you want to master on M/S? It's a recording mode to let you adjust width after recording. Is there any circumstance it would be useful as a mastering format? How many people would be able to decode it back to L/R? The other thing is that it's pointless? There's no advantage. The stereo image has been carefully tweaked and is perfect - the benefit of M/S is gone - surely? Nobody can even listen to it with de-matrixing. I guess it's a nice mono mix if you dump the side channel, but I just can't see any benefit whatsoever.

The point about using M-S in mastering is that you can treat the centre and sides separately, for example, with EQ, effects or compression. Until you have processed recordings in this way, it's difficult to appreciate what a powerful tool this method can be - if used sensitively.

Forgive the stupid questions, but I'd not realised that mastering was being done this way. I can see the benefit now explained, but is this actually common? - I assumed, obviously wrongly that the mastering engineer was always supplied with a straight left and right.

paulears, post: 446029, member: 47782 wrote: Forgive the stupid questions, but I'd not realised that mastering was being done this way. I can see the benefit now explained, but is this actually common? - I assumed, obviously wrongly that the mastering engineer was always supplied with a straight left and right.

They may well be just be supplied with L+R, but that doesn't mean they won't convert to MS if they believe they need to.

paulears, post: 446029, member: 47782 wrote: Forgive the stupid questions, but I'd not realised that mastering was being done this way. I can see the benefit now explained, but is this actually common? - I assumed, obviously wrongly that the mastering engineer was always supplied with a straight left and right.

Most of us understand M/S but might not know what mastering engineers do with this process.
http://recording.org/threads/m-s-mastering.46038/
Although I am not an official Mastering Engineer of audio, I choose to follow mastering engineers thought processing to a "T", use their choice of equipment in a M/S hardware /software matrix to mix or master all the time.
If you want to find some great info on M/S tips, there are hundreds of discussions in our mastering forum going back to the late 90's here. http://recording.org/tags/mid-side-processing/
Good quality M/S equipment allows the engineer to hard bypass dedicated tools without notice.
Being able to fine tune the mid or sides without effecting phase of a mix makes mixing extra fun and rewarding.
Miss use of M/S processing can be a bad thing as well. New arrivals to our pro audio world discovering M/S processing will usually find its first "wow" factor... when they hear the impact it has on the image of a stereo bus, later to discover they turned a once usable mix into a phased out mess.
Thus... mastering techniques can sometimes reverse bad use of it too.

• To name one M/S application on my radar: applying HPF on the sides can be a very effective way to tighten up a mix. Thus, keeping the kick/bass freq subs dead center and the sides with less sub information.

For those who do not use quality M/S hardware tools, there is excellent DAW software that imho, will M/S process better to hardware.
Hardware on the other hand can colour the mid/sides in a way digital can't.

I prefer a combination of a Dangerous Master uncoupled with a second DAW. The combination of hardware and software mid side processing is very fun.

Boswell I have one question...
you wrote:

'either export your mix as L-R and use the M-S encoder at the input of the mastering stage or export as MS'

wouldn't it have to be MS decoding at the input of the master?
it's strange because in the video nothing is mentioned about this, but obviously it works for him.
in the mixer, the db meter of my mid and side groups also differ from the video.
the mid channel only shows a signal on the left and the side channel only shows a right signal, as if both were panned hard left and right.
strange...

well, i did put the MSED on the insert of my Main LR Stereo Out of my DAW,
put it in decode mode and the stereo imaging was good again.
i guess that special hardware equipment would be needed to export as MS?
my zed r16 only has standart LR Main out, so i guess i can't export as MS
it's ok now, i can play with the mid and the side signal, which is what i wanted to do in the first place.

thanks everybody for the replies.

vintagelover, post: 446035, member: 50268 wrote: Boswell I have one question...
you wrote:

'either export your mix as L-R and use the M-S encoder at the input of the mastering stage or export as MS'

wouldn't it have to be MS decoding at the input of the master?

I'm not a full time mastering engineer but I've been drafted to do it on occasion. I would want the mix file as LR. Although I may convert it to MS at some point in the chain for processing I will probably want to process it in normal stereo first, and the mastering-specific mid-side processor I use would do both conversions to make it seamless, with the processing sandwiched between them.

That said, LR and MS contain the same information, just distributed differently. Converting between them is a fairly trivial thing, even without a special plugin. Getting the mid channel is easy, just sum left and right. You can get the side channel by summing the left channel with an polarity inverted version of the right channel. Converting back is a matter of summing the center with the side channel for left, and with a polarity inverted version of the side channel for right.

Back in the old days we would decode MS to LR by running the side channel through an XLR splitter that had one split wired pin2=cold. Pan those two inputs hard left/right and mix with the mid channel to get stereo.

bouldersound, post: 446036, member: 38959 wrote: That said, LR and MS contain the same information, just distributed differently.

Q: If the point to a mid /side process was to remove excess sub information on the sides, not effecting the mids example: (kick and bass ), are you actually removing side subs / not redistributing?

audiokid, post: 446039, member: 1 wrote: Q: If the point to a mid /side process was to remove excess sub information on the sides, not effecting the mids example: (kick and bass ), are you actually removing side subs / not redistributing?

The effect of high passing the side is to bring all lower frequencies to the center. It's almost inaudible in speakers but can improve headphone listening. It can also keep the needle in the groove.

[Edit] The only information you would lose is difference signal, things that are identical in both channels but with opposite polarity. If you have a bass instrument panned hard, the low end will still be there, just centered instead of panned. If you have two LF instrument panned hard to opposite sides you would lose some information because they would not be in phase all the time.

vintagelover, post: 446035, member: 50268 wrote: Boswell I have one question...
you wrote:

'either export your mix as L-R and use the M-S encoder at the input of the mastering stage or export as MS'

wouldn't it have to be MS decoding at the input of the master?
it's strange because in the video nothing is mentioned about this, but obviously it works for him.
in the mixer, the db meter of my mid and side groups also differ from the video.
the mid channel only shows a signal on the left and the side channel only shows a right signal, as if both were panned hard left and right.
strange...

well, i did put the MSED on the insert of my Main LR Stereo Out of my DAW,
put it in decode mode and the stereo imaging was good again.
i guess that special hardware equipment would be needed to export as MS?
my zed r16 only has standart LR Main out, so i guess i can't export as MS
it's ok now, i can play with the mid and the side signal, which is what i wanted to do in the first place.

thanks everybody for the replies.

I think Boulder has covered this, and it looks as though you have all the right procedures in place.

I would use the main L-R mix from the Zed-R16 but take it out via the Zed's FireWire interface and use that as the input into your DAW. In the DAW you should be able to flip between L-R and M-S loss-free. Go into M-S when you need to process the M and S channels differently (e.g. for centering the low end, as Boulder mentioned), and then convert back to L-R again. It's the same set of equations for both directions, the only difference being a 6dB gain change, but whether you have to account for this yourself or whether the DAW does it automatically depends on how the DAW implements the encoding.

Don't forget: always monitor in L-R. If you try listening to the M-S encode your head will unscrew.

Watched the video and realised just how useful this is - and proves you're never too old to learn - thanks guys.

yes it's a really useful and pretty logical once you get the idea.

i 'd be curious as to how hot you guys run the mid and the side channel typically when mastering in MS. both rms and peak

ok, just for the record, i know what caused my problem.
while i automatically went for encoding/decoding, the video uploader didn't bother about that at all.
he simply let's the MSED in default mode, and mutes the mid and the side signal on the respective groups...
no encoding -> no decoding!

Boswell, post: 446018, member: 29034 wrote: If you want to master in M-S, you should either export your mix as L-R and use the M-S encoder at the input of the mastering stage or else export as M-S (as you have been doing) and skip the encoding at the mastering input.

I've alway wondered why when I export with PT I get L and R files as well as the tracks. This must be why.

Hey guys, sorry I'm late to the party...
I've been introduced to M/S when listening to Fab Dupont videos and I apply that knowledge all the time
I've done a fast read of the thread and I'm a bit surprised people would go to all the trouble of splitting the signal in diffent tracks or busses when many plugins can operate in M/S.
I get that if you do, you can use any plugin, even those who don't have the M/S option... But still, in my mind, if they don't offer it, they might not have been designed for mastering.. I know I'm assuming too much here ;)

Anyway, I thought I'd join in and add that idea for those who are interested to M/S but wants to start slowly or with an easier way..
I'm using iZotope Ozone when mastering and they thought about the need for M/S work. All their modules have the option. When turned on, the module would have two sets of settings one for mid and one for side.. I find it very fast to work with.
Sometime I would turn that one not to fix a problem but just to make the song a bit wider.
I'm not saying you have to use Ozone. Many other plugins offer that option, just need to choose them carefully.
Also I do believe that we should use different plugins from the mixing phase when mastering. If using a plugin that have a certain sound when mixing adding that sound to the master as well might add up badly.. (Just a thought)

The reason I still do it manually sometimes is because I've been doing it that way since the late 80s, and copying a track seems like a trivial effort compared to soldering up XLR splitters and patching things to get that result.

If you need a hammer and the closest thing you've got is a rock, the rock is a hammer. If it gets the job done to your specs, what's the difference? "Design" is as much in the user's choice as in the maker's intent.

If I had to choose between using a separate set of mastering plugins in the same room that a song was mixed in or using the same plugins used in the mix but in a different (acoustically acceptable) room, I'd go with the different room. Even if the sonics are different, it's the convenience factor that gives mastering-specific plugins their main advantage. A limiter designed to give you volume without overs is just way more convenient than a generic limiter, so the time and energy you save on that process can go toward making things sound better.