Mid-side decoding issue - Total silence from the two side tracks

[quesne]

Hi Folks:

The attached zip file contains my original, manual attempt at M-S decoding (i.e. no plug-ins, routing or FX). I simply recorded my side-shifting vocal using the two mics (CAD M179 on Track #2, and Spark Blue on Track #3). I then copied the CAD audio to a third track (Track #4), and reversed it's polarity. Finally, I hard panned the CAD tracks left (Track #3) and right (Track #4). The raw WAV files are the exact same ones I put up on Dropbox the other day for everyone to download. There are no plug-ins or routing changes to the tracks, and no exotic settings (that I am aware of) in the project.

For me, this project does not exhibit any L-R stereo separation at all. As we now know from our combined analysis, there is nothing wrong with the audio. Everyone -- including myself -- has gotten them to work correctly. So, almost by definition, the problem is either in the project itself, or in my Reaper settings. The other possibility, of course, is that I am missing something so glaringly obvious that someone will let me know in the first 30 seconds! But if it's there, I don't see it yet.

Let me know if I need to provide any other info

 

[quesne]

FYI Papanate - With respect to the screenshots, I just had a look, and there are 46 separate screens (i.e. categories of settings) in the Reaper menu system! So I'm not really sure where to start On top of that, there are also Item Settings, Item Properties and Source Properties for each of the audio files.

If anyone out there does have Reaper -- and has some idea which config screens I should post -- please let me know

 

[Steenamaroo]

You've got everything going through a master track/bus, with centre panning.

I have a ProTools skin on Reaper so I don't know if the layout will be the same for you,
but your master should be panned hard left for left and hard right for right.

 

[Steenamaroo]

Ok, I flicked back to default theme.
Right click on the pan knob at the top of your master track.

Change pan mode to Dual Pan.
Ensure the sliders are dragged hard left and hard right (they should be already)

Alternatively you can set pan mode to Stereo Pan then increased the (newly revealed) width control to 100%.

Either should sort it out.

 

[quesne]

OMG! Yes, when I change the pan law of the master fader from "Project Default (Stereo Balance)" to either "Stereo Pan" or "Dual Pan" (see attached), it corrects the problem It's awesome to have a definitive answer to the "How" question! Thanks Steenamaroo

The "Why" question still seems to be hanging out there, though. The weird thing is that, with every other stereo recording setup I have used in the past (including the Voxengo MSED solution mentioned previously), the "pan law" of the Master track never came into play. Even though the decoded M-S signal is passing through the exact same Master fader when it's set to "Project Default (Stereo Balance)".....it has always worked fine

So what do you think is going on here? Does the decoded M-S signal simply baffle the "Project Default (Stereo Balance)" setting to the point where it needs to be "forced" into a correct separation via the "Stereo Pan" setting? Or is there a setting somewhere in the Reaper defaults that actually changes the behavior of the "Project Default (Stereo Balance)" setting behind the scenes? This is a real mystery.

For no good reason, I had been putting off upgrading my Reaper install. I am going to do that ASAP and see if my other past projects behave differently. Thanks again Steenamaroo! And thank you to all who have been participating in this exercise It has been educational

 

[quesne]

FYI - Just to add one more important data point to the analysis....the attached project (created by one of the Reaper forum guys, and using the same exact audio files) does NOT have the Master fader pan law set any differently. It is on "Project Default (Stereo Balance)". Yet it sounds and works perfectly for me! So what is the difference?

Not sure why, but it wouldn't let me upload it. So I put it into DropBox:

Dropbox

www.dropbox.com

 

[PorterhouseMusic]

 

[Steenamaroo]

It seems I was right, but not quite right.
The problem wasn't your choice of setting for the master track output - It was the fact that width was set to zero.

Changing the setting exposed the width control, allowing you to turn it up and solve the issue,
but you can switch back now (to Stereo balance / mono pan) and it will still work.

That explains the project you received working fine. Whether visible or not, width must have been set to 100%.

Not sure where else width can be set - I don't use Reaper.
I suppose it's possible there's a shortcut that was input accidentally or something like that?

 

[quesne]

Yep, I can see that now Steenamaroo.

FYI - I found the spot where you can change the project default for the pan law (see attached images). It's a little weird because, even after you change the pan law for the current project -- and then make it the new default for all new projects -- the Master fader UI on new projects does not change even though it's behavior changes behind the scenes

The individual track UIs change (the knobs at the bottom left in the image are "pan" and "width"). You can see on the track that the width is 100% by default, but it's not so obvious on the Master fader. When you left click on the Master fader pan knob, however (just like Steenamaroo points out in his previous post), you can see clearly that the width knob is at 100% by default! So the behavior of the Master fader does change in the backend, even though the UI control does not.

Anyway, the full explanation for what's been going on in my Reaper DAW -- and the solution moving forward -- are both pretty clear (i.e. the "why", "how" and even the "where" This has all been an instructive and useful exercise. I hope it is also helpful to someone else in the future!

 

[Steenamaroo]

I think, although welcome correction, pan law governs how loud audio is the farther away from centre it is panned.
Ordinarily if you panned slowly something from centre to one side it would seem to get quieter.
I believe you can set custom pan law to compensate for that.

Regardless, as you say, we have the why, how, and where, so I can get back to my Ledaig Rioja!

 

[PorterhouseMusic]

I think, although welcome correction, pan law governs how loud audio is the farther away from centre it is panned.
Ordinarily if you panned slowly something from centre to one side it would seem to get quieter.
I believe you can set custom pan law to compensate for that.

Regardless, as you say, we have the why, how, and where, so I can get back to my Ledaig Rioja!

Cheers. I'll drink to that.

 

[TalismanRich]

I just finished all the yard work, and checked out the thread to see if there was any resolution. I pulled down the file, and my initial thought was that you were sending the track to a "mono" subgroup track where you can only pan left/right. That didn't turn out to be the case.
It was interesting that after I tried loading your RPP file, I got the same result as you. The issue was the width control of the pan. Somehow it was setting to 0.

The folder I downloaded from Dropbox is essentially what I did (except for how I did the invert).

Here is my complete project, with the same subgrouping that you did. If it doesn't pan the sound, make sure that your pan widths are set to 100.

MS pan w/ subgroups

 

[quesne]

I just finished all the yard work, and checked out the thread to see if there was any resolution. I pulled down the file, and my initial thought was that you were sending the track to a "mono" subgroup track where you can only pan left/right. That didn't turn out to be the case.
It was interesting that after I tried loading your RPP file, I got the same result as you. The issue was the width control of the pan. Somehow it was setting to 0.

The folder I downloaded from Dropbox is essentially what I did (except for how I did the invert).

Here is my complete project, with the same subgrouping that you did. If it doesn't pan the sound, make sure that your pan widths are set to 100.

MS pan w/ subgroups

Hey TalismanRich:

Yes, absolutely As long as the pan law on the Master fader is set correctly -- allowing me to set the width to 100% -- your project plays back flawlessly

 

[DrewPeterson7]

Hmmmm....I have not yet experienced that sort of sonic "instability" on my M-S recordings (at least not the ones that worked well But I must admit that I rarely listen to my mixes without headphones. I am going to go back and test a few of my good mixes on the studio monitors, and see what happens when I shift my ears relative to the speakers.

I would STRONGLY recommend listening to a mix through real speakers before finalizing it, if you're doing a lot with mid-side techniques.

The problem with mid-side and headphones, is the out-of-phase tracks never get a chance to interact - one goes straight from your left headphone speaker to your left ear, and the other goes straight from your right headphone speaker to your right ear, and your head basically stops the waveforms from ever meeting.

In a room, when you play them back, there's no "barrier" like that. They won't entirely cancel to zero thanks to room reflections and differing distances from one speaker to the other, but that's what's producing that weird hollow sound in the room, very heavy phase cancellation.

The upshot is, phase isn't a huge deal for a headphone mix, because unless tracks are panned together, you'll never actually hear phase cancellation since the audio doesn't ever "meet." It's like Ghostbusters and crossing streams. Never cross streams. And, when you take that mix and play it back into the same acoustic space, you WILL cross streams. Phase cancellation will make mix translation an absolute nightmare - mixes will sound way different on headphones than they will over speakers.

Seriously, take one of these mixes you're pretty happy with, and find a good stereo speaker system and play it back, and just walk around the room while it plays. See if you even recognize it.

 

[quesne]

I would STRONGLY recommend listening to a mix through real speakers before finalizing it, if you're doing a lot with mid-side techniques.

Thank you for pointing this out, Drew. I had not really thought about the lack of L-R wave interaction when listening through headphones. I will try to always make a habit of listening to my mixes through my studio monitors before they get finalized

 

[TalismanRich]

I don't think there is any more of an issue with M/S vs any other stereo technique. All the summing and difference manipulation is done in the electronic realm, not magically in the speakers.

Checking on speakers goes for ANY mixing project. I will listen on headphones, but any decisions on placement or adjustments to the sound come via monitors.

 

[bouldersoundguy]

I suspect simply pressing the mono button would reveal phase or polarity problems between L and R. Just remember to switch it off when you're done. Also, if you check it on a surround system and some things appear only in the surround speakers (and it wasn't intentional), that's another tell.

 

[TalismanRich]

One of the advantages of M/S is good mono compatibility.

The other advantage is that you can adjust the "spaciousness" of the stereo image by adjusting the relationship between the M and S channels. That was evident in the difference between Porterhouse's first mix and my first mix. As you blend then two, the strength of the side channels changes. You can't do that with an X/Y or other 2 mic techniques.

 

[DrewPeterson7]

One of the advantages of M/S is good mono compatibility.

The other advantage is that you can adjust the "spaciousness" of the stereo image by adjusting the relationship between the M and S channels. That was evident in the difference between Porterhouse's first mix and my first mix. As you blend then two, the strength of the side channels changes. You can't do that with an X/Y or other 2 mic techniques.

Then I must have done something wrong when I gave this a shot. My workflow was this:

*Position a condenser in cardioid mode facing the source (here an acoustic guitar), with a ribbon aligned 90 degrees above it, so the figure 8 pattern was side to side.
*record to seperate tracks, then duplicate the ribbon track, pan each hard L and R, and flip one of them out of phase. Bus together in a stereo bus. Start with bus volume all the way down.
*pan the condenser to the center, then gradually pull up the bus volume to widen the stereo spectrum.

It worked... but I also heard major phase differences as I moved my head around. Does this require some special way to record the two sides of the ribbon separately, or would a pair of cardioids back to back work better?

 

[Steenamaroo]

It's such a great technique for a few reasons.
Of course it's very easy to adjust the perceived width but you don't have the concern of distances between microphones
or between each microphone and any given source,
so it collapses to mono perfectly and doesn't suffer from potential phase issues.

If you're having phase issues with mid/side you're doing it very wrong.
The microphones should be about as close together as is possible.