Do any of you guys here experiment with M-S processing in a mastering situation and if so what type of songs benefit from such techniques.

Yes I do. It's used mostly in songs that are not mixed correctly or if a client wants a particular voice emphasized. Sometimes it can be done. It's great for increasing/decreasing the depth/width of the sound stage where the mixing engineer screwed up.

Regards,

or in the case of some, ending your career

Spill the beans then. Who / which tracks are you thinking of?

Anyone care to elaborate on how to set this up?

Thanks

2 channels of Left and Right stereo information can be easily converted to 2 channels of Mid and Side information, and then converted back again, with no losses in stereo information.

The advantage of converting to Mid and Side is that the balance between them can be altered, or you can process one without affecting the other (if you're careful about phase).

The maths is basically this:

Mid = (Left + Right)
Side = (Left - Right)

Mid is just a the mono'd Left & Right (compensate levels).

Side is a mono track, representing the information going on at both of the sides. Obviously centred audio information, like a dry vocal, will disappear because they are the same on both sides. Doesn't actually matter whether it's left-right or right-left - the difference is the difference, and only the polarity changes.

(Side information can be recorded at the tracking stage with a mono figure 8 mic aimed at both sides).

Encoding Side from a Stereo Left/Right track is your basic Karaoke vocal removal trick.

You achieve this by assigning Left and Right to two seperate mono channels in a group. Sum them and drop the group fader 6dB to get Mid. Or Sum them, invert the phase on one side and boost the group fader 6dB to get Side.

Then to convert back to Left & Right, do this:

Left = Mid + Side
Right = Mid - Side

Again, using your mixer and phase switch as required.

Or use a MS plugin for encoding/decoding if you have one available.

So if you alter the level of the Side channel before decoding, you can effectively pan backwards and forwards, instead of left to right. Or in otherwords, from stereo wet to mono dry. You can boost the amount of Side information, effecting increasing the reverb by boosting the ambiance that was already there.

Or using phase accurate digital effects, you can alter Side information without effecting Mid, or vice versa.

Edit: I just took out any references to boosting or cutting by 6dB. Obviously when summing similar tracks to mono, you get about a 6dB boost. Just adjust levels as necessary - the main thing is the Sum and Difference aspect, the use of mixing and phase reversal where needed.

Very interesting...I'll have to give that a shot in Logic.

Thanks for taking the time to explain!

It's saved my a$$ more times than I can count. Buried vocals, overcooked ultra-panned guitars, center EQ disjointed from the sides... On and on.

very useful.........i use it a lot......one of the ME's main WOMD!

LINK to more info

...and watch your levels in the chain.

logics got a plug to do it called "direction mixer" (I think that's what its for, never used it)

narco

You can widen a mix with just four faders. Double the mix up on four faders. One pair standard pan left and right. The other pair reversed pan and reversed polarity (on both faders of the second pair). As you add in the second pair the total level will GO DOWN and the width will increase. So you have to make up for it by adding level.

The Waves S1 accomplishes the level makeup internally but uses the same technique. So the width control is just width without any muss or fuss.

But you can't do frequency-sensitive M-S processing with the S-1. By using the four-fader technique, and putting filters on the second pair of faders you can limit the frequency ranges in which you widen.

But WATCH OUT WATCH OUT. M-S processing changes the mix. 100% it changes the mix. A mastering engineer has to be extremely sensitive to the balance that the mix engineer was trying to achieve, and not suddenly focus on the "oh boy, it's wider" effect!

Warning, Plug coming from product designer. WARNING. PLUG COMING!!!! That's why I invented the K-Stereo. It has a depth-increasing, ambience-increasing and width-increasing effect with minimal effect on the original mix. But it should be used in a mastering context, it is really not a mixing tool.

K-Stereo also includes an M-S control so if, for example, you have a vocal-light mix and a remix is not possible, and you have no choice but to raise the M channel, a little of the K-stereo effect added can restore some of the lost ambience, width and depth that you lost by raising the M channel.

Hey Bob,

I can't resist. I played with the k-stereo on the Weiss unit at aes and was really impressed. When will we see plug-in versions? Now that Waves and Audioease are both cross-platform and ilok (and even PSP is going that way now), do you feel like you'd have the licensing security to do a plug-in?

best,

r.

Hi, Rob. Thanks for your comments. I didn't think you could hear anything with earphones at the AES show. I've been resisting a plugin for a long time... Mastering is a small, niche market as it is. Costs versus return, and protecting the hardware vendor(s).

Probably not, but something could make me change my mind.

BK

Pleeeeeeeeaaaaaaaaaaaaaase Bob
for pingu

I'm sure I wasn't able to get the full effect, but the acoustic guitar demo in particular really knocked my socks off. I was really struck by way that your application of stereoization is to augment and enhance that elusive space around an object, rather than gimmicky, dual mono, speading of left and right. I really loved how it pulled the natural ambience out without making me feel like I was paying a deep price of phase sickness.

I don't know how much of a licensing fee you can be making off each unit of hardware being sold and how many units can realistically be sold, but I'm sure that you are seriously underestimating the number of pieces of software you could sell if you get the pricing right.

I won't twist your arm anymore, but I suspect you have a much more substantial product than you think.

best,

r.

M/S is a powerful tool that can be used for good or evil. It's one of those things that's great to pulll out when there's a problem and it can seem like a magical fix. However, when used as a matter of course, it can weaken the center, the impact, and certainly the balances and imaging that the mixer worked hard to create. I think it's more of a problem fixer than a daily workflow tool. Like other powerful mastering tools, don't grab for it just because it's there. Grab it when you specifically need something it offers.

It's not just limited to EQ. Dynamics processing in M/S can also be very useful. The days of vinyl saw some of this (think vertical and lateral).

De-essing in M/S is also a good application. The "SSS" sound on the vocal is most likely up the center, so you can de-ess without hearing your your cymbals and stereo guitars dipping every time there's a bad "SSS" getting caught.

There are usually some multiplies by 0.707 in that path.

DC

You can use the S-1 for this purpose, provided you possess an EQ plugin (such as the Waves Q10), which allows for separate adjustment of left and right channels.

The technique is explained on page 16 of the S1 manual (S1.pdf) which can be downloaded here: http://www.waves.com/downloads/pdf/Diamond.zip

ps hello bob!

How would you suggest bringing out a slightly buried vocal once the mid channel is seperated? This would save me a lot of time, otherwise I would have to go back in to the original mix and fix it there.

Cheers

Mike

the TC MD 3 has a setting for this

That would be nice ($$$!) but we have the Waves stuff.

Cheers

Mike

Hello NIGEL!

Right... well, in that case you might as well skip the S1 and just use the simpler Waves MS encoder in front of the EQ and decoder after it. Most of the tools that allow for separate M and S channel processing require putting the encoder in front of a dual-channel processor. There is no "insert point" in the S1 per se that would allow you to adjust the width control on, for example, "just the midrange". So the controls in the S1 become less powerful and most of the adjustments you will be making will be in the other processors and so the S1 becomes relegated to mostly an encoder/decoder, wasting CPU cycles.

However, the Gerzon/Blumlein style manipulations in the full S1 are very useful on some material, and those are frequency-sensitive.

BK

Not really, it generally sounded like crap! Digital M/S is much better sounding but that only makes it more dangerous than when it was a no-brainer to hear what that switch on a 670 did.

The whole problem is that we'll always tend to hear what we expect to hear without noticing the downside unless we've learned to listen carefully for that in addition. In my experience the more hype there is around a device, the harder it becomes for people to hear the problems with it. I've often thought the only secret of recording may be learning to manage one's brain.

What do you mean by "not really?" There were several mastering pieces around that offered it, both EQ and dynamics. I never said it sounded good or was used often, just that it's not new. Some young people get a new plug-in and think the technique has never been thought of before.

I meant it wasn't used.

M/S dates back to the 1930s.

I'm using a Sony ECM-MS5 mic for some tracking. I must say I like the performance of this mic and all the wiring and conversion is taken care of. It adds that little 'edge' to otherwise mono instruments. The only problem is in mixing when I now have to deal with two signals instead of one when cleaning up the tracks.

Brilliant. Hope you don't mind that I nicked it.

The Phoenix "Nicerizer" 16 has a width control which I have yet to actually use on a full mix. The claim is that it doesn't mess with the low end, but it seems to spread it which means a loss of punchiness. I love this box for summing, but I wonder if anyone is using this width feature. So far I used it make a stripe of BGV to add to the original tracks. We were going for many voices singing the final chorus and it was effective in that role. I've asked Shaun what exactly it does, but haven't gotten an answer.

Neale Eckstein
www.foxrun.org