Skip to main content

I'm struggling with the age-old problem of getting exactly the right mix of the lead vocal in relation to the instruments.

I've noticed that when I'm trying to preview my mix on different types of speaker and headphone systems, I find that the relative level of the lead vocal compared to the instruments changes, depending on the amount of stereo separation in the playback equipment. I'm looking for feedback on how other producers handle this particular issue.

I find that as stereo separation increases, the lead vocal seems quieter, and as stereo separation decreases, the lead vocal seems louder.

I believe it's because, in general, instruments are traditionally recorded in stereo, while lead vocals are recorded in mono and panned center, and we're getting additive phase for the lead vocal because of this.

I believe that when I play back the mix on stereo speakers, I am getting the effect of additive phase; the actual sound of the lead vocal (at least the main non-reverb'd part of it) is identical coming from the left and the right speaker, so each wave is adding to its twin as the Wavesmove through the air in the room. The closer the speakers are placed to each other, the greater the effect.

The opposite effect happens when I listen to the mix on a pair of headphones, which can be said to have 100 percent stereo separation. There is no additive phase because each speaker sends a unique signal to each of my ears. The fact that the two signals happen to be the same isn't relevant and isn't interpreted by my brain as increased volume.

On the other hand, stereo-recorded instruments will be sending somewhat different sounds to each speaker in terms of the timing of each of the waveform peaks, and so there is much less additive phase happening. Thus, on a pair of headphones, they seem louder in relation to the vocals.

The result is: Listen on speakers=louder lead vocal, listen on headphones=louder instruments.

I know that the mantra is not to mix for headphones, but in this Age Of The Ipod, I know that a significant portion of my albums' listeners will be listening on headphones (including those terrible white Apple earbuds, ick), so I don't want to discount them.

What I've tried so far is to add a stereo simulator to the lead vocal (in effect just combing the frequencies of the mono signal and panning them LR), and this helps a little, but strangely, it doesn't help as much as I thought it would. If my theory were correct, it should have completely fixed the problem.

My next step is to try recording vocals in actual stereo, too, but I worry about that.

Has anyone else run into this, and what do *you* do about it?


anonymous Tue, 08/18/2009 - 15:14

If I sum the whole mix to mono and listen on my monitor speakers, I get the expected result based on my theory: It behaves the same as if I were standing back from a pair of stereo speakers that were right next to each other with no stereo separation. In other words, the vocals seem louder in relation to the instruments than they were with the speakers in a nice wide stereo pair. And of course, summed mono played over a pair of speakers sounds the same as summed mono played over a pair of headphones.

TheJackAttack Tue, 08/18/2009 - 15:18

You could try duplicating the vocal track. Leave the original the way you currently have it mixed and then try your stereo simulator on the duped track.

As an alternative, maybe there is too much separation on your stereo instrumental tracks. As a live engineer I am rarely recording anything but the main pair in stereo.

Caveat: I am not a mastering engineer so I definitely do not have all or even most of the answers.

anonymous Wed, 08/19/2009 - 12:53

The perceived volume level of the instruments does not change when I mute the vocal. The perceived volume level of the vocal does not change when I solo it.

It's just that the perceived volume relationship of the vocals to the instruments is different depending on the amount of stereo separation in the playback equipment.

Is that what you were asking?

cfaalm Wed, 08/19/2009 - 14:46

Not exaclty, but maybe my question was inadequate.

What I was after is what happens in these situations?:
A. just the vocal on speakers or headphones in stereo
B. just the vocal on speakers or headphones in mono
C. just the instruments on speakers or headphone in stereo
D. just the instruments on speakers or headphone in mono

I am trying to see if this can be isolated. If not then my next guess would be the vocals and instruments are send to the same stereo effect which

Then again, I'm not an expert. Hope a real expert can chime in at this point.

anonymous Wed, 08/19/2009 - 14:54

See, that's just it: "Just the vocals" or "Just the instruments" sounds fine in all cases. It's only the perceived difference between the two that's at issue.

Muting or soloing one or the other just makes that element disappear, and doesn't change the other elements which remain.

The instruments and the vocals aren't being sent to the same effects, or anything similar that might otherwise cause a phase issue. All instruments and vocals have their own effect inserts, and besides, this happens without effects, too.

I'm talking about a general overall thing I've noticed in all mixes, not just one particular technical problem. Has anyone else noticed this kind of thing happening, or even checked to see if it happens when listening yourself?

anonymous Thu, 08/20/2009 - 07:32

The solution to this should not be a complicated one.
Resorting to any unnatural processing usually makes matters worse.

I've had this problem before and I fixed it by changing my mixing technique.
I used to mix at very loud monitor levels.

I think you should try mixing at very low volumes.
Get the vocal sitting right there and it should sound right everywhere.
I'm almost certain you're mixing at loud or too loud levels.

anonymous Thu, 08/20/2009 - 10:54

I definitely mix, and preview my mixes, at both loud and soft volume levels, and I preview my mixes on different types of playback equipment. I agree this is an important part of mixing.

Issue is still the same at any volume level: Increase the stereo separation in the playback equipment, and the relative volume of the mono-centered vocal compared to the stereo-recorded instruments changes.

cfaalm Thu, 08/20/2009 - 12:17

I have one other option in mind: The mix is too stereo. :o :?
But I really don't know if it this is true :oops: . Please, someone slap me if this is incorrect:

What stereo mic technique was used for the stereo instruments?
Are all of them panned hard left and right?

My point being the stereo spread could be too wide, wider than the real thing, which thins it out. With the monitors close to eachother this isn't obvious, as it is almost a mono mix you're listening too. With the monitors further apart the spread in the mix is directly translated, exaggerated even when they are further apart then the instrument sounded in front of the microphones.

So perhaps narrowing a couple of stereo spreads down could make less instruments thin out as the stereo spread gets wider. If you feel you're losing scope you can fill the outer ends of the mix with mono details like percussion, ad libs, fly by sounds that appear only scarcely. Even monofying a few might help.

In other words, rethink your stereo image from instrument to instrument.

anonymous Thu, 08/20/2009 - 12:44

But I lurve my stereo instruments. :-)

I agree, though, that this sounds like a valid possible solution to the problem. Is it common in professional recordings to deliberately make the stereo spread smaller? Seems like a step backwards to me.

To answer your specific question, my instruments are mostly recorded as XY stereo, or Mid/Side stereo.

TheJackAttack Thu, 08/20/2009 - 12:57

A stereo image is not always spread as wide as possible. Often that is not the natural way we perceive sound. Of course "stereo" itself is a manipulated concept too but I digress. You can rarely get too much spread with a proper XY pattern but for M/S it can be very easy to overpower the mid. I use M/S frequently in live concert recordings specifically so I CAN manipulate the spread easier, and often I do have it "smaller" as you put it. I would go so far as to say that for commercial recordings, most instruments are not even tracked in stereo at all. By the time you have everything duplicated and manipulated it is terribly difficult to minimize phasing and harmonic buildup if everything is tracked in stereo.

Again, YMMV and mine is not the only way.

RemyRAD Thu, 08/20/2009 - 16:05

You are talking about a number of different situations You can't lump them all together. So your recorded vocal was recorded with a single microphone? Stereo microphone? Items in appearing equally in both channels that are common to one another will appear 6 DB louder when collapsed to Mono. Similarly, speakers will exhibit similar phenomena because signals that are in phase and identical will increase the center image as they get closer so that their energies can be combined, acoustically. There. Wasn't that simple? And so now you understand how to interpret everything you hear. Nice thread.

I need some new clothes
Mx. Remy Ann David

anonymous Thu, 08/20/2009 - 16:59

Items in appearing equally in both channels that are common to one another will appear 6 DB louder when collapsed to Mono. Similarly, speakers will exhibit similar phenomena...

Aha! Yes, exactly! That's what I'm talking about.

So. Knowing this...

Let's assume that I *want* to record and mix my instruments in full stereo, and I *want* to record my vocal in mono and pan it center. Furthermore, let's assume I want my listeners to be able to get approximately the same amount of relative lead vocal, whether they're enjoying my music in mono, on poor stereo speakers, on nice stereo speakers, or on headphones.

What are my options for improving the mix in this situation, in such a way that it mitigates this 6db difference?

I realize that I'm asking the impossible. The things that I want are at cross purposes from each other. However, we've got 21st century technology, and over a hundred years of collective historical audio-recording know-how, at our disposal. Plus, I'm sure the "pros" already have a solution to this, and I'm really dying to know what it is.

The options I've seen so far (including the good suggestions in this thread) are:

- Add a stereo simulator effect (splitting frequency combs L+R) to the lead vocal.
This helps, but not much. Not sure why.

- Bring the instruments down to mono, or close to mono.
Ewwww. Ick. I spent all this time getting a gorgeous stereo image out of the instruments, and you want me to throw all that away? Besides, I listen to albums by the "pros" and their instruments all have great stereo images.

- Record the vocal in stereo.
Seriously considering this one. But because of how close-mic'ed the vocals need to be, I'm worried the vocals will move around in the stereo field too much. And of course, summing it to mono to force it to the center would defeat the purpose and we're back to square one. Still, does anyone have any successes to report with this?

- Give up, and choose to mix just for my nice stereo speakers, and screw the ipod listeners (at one end of the spectrum) and screw the mono listeners (at the other end of the spectrum).
I'd prefer not to do that.

- Pretend I'm The Beatles and pan the vocals hard to one side.
I'm not good enough to make that work for me. :-)

Any other ideas or experiences that might help?

cfaalm Fri, 08/21/2009 - 15:25

tfabris wrote:
- Bring the instruments down to mono, or close to mono.
Ewwww. Ick. I spent all this time getting a gorgeous stereo image out of the instruments, and you want me to throw all that away? Besides, I listen to albums by the "pros" and their instruments all have great stereo images.

I didn't say all the instruments. Depending on which instruments you have you could study the arrangement and see which instruments need to be full stereo, which ones could do with less, which ones could be mono.

If you do record the vocal in stereo, I suggest you use MS. It will provide a stable (stereowise) signal in both mono and stereo.

pmolsonmus Sun, 08/23/2009 - 06:32

Maybe this is obvious, but you should be checking your mix in mono throughout the process. And, as a vocalist and teacher, the bottom line is that you should make the vocals work because that's what 98% of your listeners are focusing on. (Sorry instrumentalists) And then, as you'll learn, it is all about compromise to make the mix work. If you want it to sound great in all (or almost all) settings, then create a great neutral mix and have a qualified mastering engineer go at it. There is a limit to DIY and there is a reason they have great rooms and great equipment.


RemyRAD Tue, 08/25/2009 - 02:04

I think if you just recorded them with a XY stereo microphone and strap their head down so they can't move their head left or right that should solve the problem.

Or just give them a pair of those over the ear microphones that run down the side of your cheek. You know, the ones you see the sportscasters using at the live games. Use one on each side. Then you won't have any problems with 6 DB center image build up problems. Yeah, that's the ticket. No phase shift. No build up. Consistent mix levels. Sounds great. Less filling. Sounds great. Less filling. And they can move their heads back and forth with out causing you any stereo problems.

I'm a genius.
Mx. Remy Ann David