Singing With Headphones

Jun 11 2021

This is a problem that has plagued me ever since I started recording myself (as professionally as possible): Whenever I sing into a microphone, with headphones on, I’m always off-pitch.

In this article, I’d like to explain the problem, reasons why it might occur, and possible ways to solve it. (Because, when researching, I found many others having similar issues singing with headphones. Hopefully this helps!)

The problem

For years, before I bought a proper microphone and audio interface, I only recorded myself in a “natural setting”. I played my guitar, sang over it at the same time, and recorded it behind my desk (with my computer microphone). These recordings sounded fine! Sure, I didn’t love the sound of my own voice at first, and I took some singing lessons, but no major issues with pitch.

Then I bought proper equipment. I recorded a piano (solo). Then I recorded my voice (solo, using the piano as a backing track through my headphones).

I remember being so disappointed when I listened to the result.

It sounded dull. Lifeless. Not like my voice at all. The voice didn’t match the piano, creating a very ugly version of the song.

At first, I thought I just had the sing more energetically. Be more enthusiastic! Put more emotion into it! But that wasn’t true. It just made the recording louder, and my voice tired.

Then I thought I had to use some mixing tricks, some plugins and effects, to blend the piano and the vocals. But no, that wasn’t the solution either.

Finally, I thought: “screw this, let’s just throw autotune on the vocal and see what happens” You guessed it: not only did it sound much better with autotune (even though I absolutely hate that sound and would never use it in a final song), it also revealed that every single note I sung was off-pitch

Not just that, all notes were consistently flat by the same amount!

Using a trial version of Melodyne, I simply shifted everything a quarter note upwards. (So, that’s roughly 50ct, or half a half tone.) And voila: the whole vocal was completely on-pitch. It sounded right. It fit with the piano, it felt energetic and powerful, awesome!

Solved, right?

Well, no.

First of all, I don’t want to rely on plugins and autotuning. I strive to be the best musician I can be, and consistently singing flat (and requiring effects after each take) is not part of that. In my eyes, this should be the case for every musician.

But secondly, autotuning brings artefacts. It introduces a robotic sound, sometimes even pops, clicks and other minor distortions (if the autotuning algorithm can’t quite make sense of what you were singing). And unless you want that exact sound for this exact song, it will sound bad and wrong and stupid.

Preferably, I’d sing on-pitch by default and forget about all this tuning magic.

Remember that I did not have this problem when singing any other way! Whenever I have a song idea, I usually record a quick demo. I grab my guitar (or race to the piano), start recording on my phone or laptop, and quickly lay down the idea. And in all of these demos, I am not flat (well, 99% of the time).

So the question becomes: Why does singing into a mic, with headphones, make me consistently flat? and especially How can we fix this?

Reason 1: The microphone

Unless you’ve done it before (countless times), singing into a microphone and recording yourself is not natural to most people.

Usually, you will adjust in certain ways:

I’ve done all of these. I might still do, from time to time, when I’m not feeling great or have an especially difficult part to sing.

But all of these things cause tension, which is the number one reason for losing control over your voice.

Reason 2: The headphones

We (humans) need feedback to constantly attune ourselves to the surroundings. Most people cannot sing the correct note, unless they can hear themselves and make minor adjustments if they hear something is wrong with their pitch.

When you put on headphones, you stop hearing yourself naturally (especially if they are high-quality closed headphones, and the music is loud), making it impossible to be certain about your pitch.

This is why you often see artists with earplugs in live settings … which they take out, then plug in, then take out again repeatedly during the concert. They need to hear themselves well, to sing well. But if they only hear themselves, or it’s too loud, they take it out to prevent confusing themselves.

This is exactly what happens in the studio.

Again, unless you’ve practiced repeatedly with headphones on, it will be an unnatural situation for you and your voice cannot find a good pitch anchor to hold onto. These situations might happen:

This caused me to consistently be flat by 1/4 note. Because when I say consistent, I mean consistent: shifting the whole track 1/4 note upwards, without changing anything else, would completely solve 99% of pitch issues.

Clearly, my voice knew how to sing the song, and how to reproduce the melody on the pitches I heard. The problem was … that I heard the wrong pitches.

Doing some Experiments

I’ve been interested in singing and vocalizing since a very young age, which means I wanted to experiment with this phenomenon, instead of just searching for a trick to fix it.

Could I change something in the way I sang? Could I do some exercise or warm-up with headphones on, that would tune me properly?

Well … yes and no.

Of course, by singing with headphones on, I get more comfortable in those situations and can compensate better.

Similarly, I can change my technique. We automatically sing “sharper” if we smile and when we bring the sound more “forward” (like it resonates around our nose and eyes, instead of our throat). I also tried singing into my microphone with a tuner plugin active, and tried to perfectly match the pitches. Both adjustments were doable with some trial and error and actually solved large parts of the issue.

But do I really want to change my technique and “relearn” singing, just for studio recording situations? Isn’t it stupid to be smiling like a manic during every take, just to sing on pitch? What if I don’t like that more “forward” sound, and would prefer my usual deeper timbre for certain parts? No, this feels like a short-term panic solution.

The fact that my ears don’t hear the pitches properly doesn’t change, which is really most of the issue.

Solutions

Which brings me to possible solutions. Hopefully, with the detailed explanation above, you already have a strong sense of what to look out for. But here I’ll present a list of things I tried and my thoughts about them.

Obviously, make sure everything is in tune

I made this mistake when I just started recording. I thought “hey, the guitar sounds fine, let’s record it!” The guitar did not sound fine. It was not in tune, and the strings were old and dull. If that’s the case, you have no hope of getting a nice vocal on top of that.)

Singing with one can off.

For example: the left can of the headphone is on my ear, but the other is tightly sealed against the back of my head.

Yes, this works quite well! You can hear yourself better, it feels more natural, but (if done correctly) no “bleed” from the headphone should enter the mic.

Turning down the volume

I’ve always listened to things on low volume. So when I did this, I found myself straining to hear the actual accompaniment! Which made me uncertain about the pitches and timing, leading to terrible performances. But if you’re a “loud listener”, this might help.

Changing the “cue mix”.

Usually, the sound that accompanies you during a recording is called the cue mix. At first, I just left the whole mix enabled and sang over that. But that didn’t work for me! Most tracks had one or multiple guitars, and a metronome for timing, and it was just messy to sing over that.

Then I started creating an actual cue mix. For me, I’ve found that I prefer the sound of a digital piano. I would grab my plastic keyboard and play the chords for the song into a standard sample library. Then I would mute everything but this piano recording, and sing over that. Most of the time, this made me sing on-pitch.

I also like replacing the metronome with an actual drum sound, or some other actual instrument to indicate tempo and groove. However, if your final song does not have drums (like with my small singer-songwriter melodies), I don’t like this, because it gives off the wrong vibe and makes you perform to an instrument that isn’t even in the final song! (Causing people to wonder why on earth the singer sounds so happy and upbeat during that sad piano ballad :p)

Yes, this can be quite some extra work, especially in busy mixes. You need to mute everything you don’t need, and only keep one or two instruments for a clear pitch and tempo. To me, however, it seems a “wonder fix”: pick the right cue mix, and suddenly your vocals start to shine.

Remove the effects … or not

I’ve always added effects after I’ve recorded everything. So for me, the cue mix was actually too “dry”. I’ve found that I actually sing a bit better if I add some subtle reverb to the vocal track. It sounds more natural. You become more confident in your voice, because it sounds cooler.

But if you’re someone who automatically throws effects on everything, you might want to dial that back. As those, too, will mess with your pitch perception. (And might cause extra delay.)

Use different headphones

The best headphones for recording are, usually, closed-back headphones of a good enough quality (so that they have enough bass response and snap snugly around your ears).

However … those are the exact things that cause our perception to be altered. They are closed, so they completely cut us off from the surroundings and directly pump sound into our eardrums. They provide a good bass response, which will throw us off, as we need mid-range frequencies to anchor ourselves.

If any of the above doesn’t work, you can try recording with cheaper earbuds, or with open headphones. If you choose to do so, however, make sure they don’t bleed into the microphone too much. (As the name suggests, an open headphone will allow the sound to escape from the other side as well.)

Update! I’ve been able to try this extensively and found some interesting results. Using very cheap earbuds, I was able to get better vocal results than my quality headphones. There was no bleed, unless you set the volume absurdly high, or sing like you want to eat the microphone. At the same time, I could hear myself transparently, and could move around more freely. It did not completely solve this issue, but it was a much nicer experience than heavy closed-back headphones.

Use the monitor-canceling-trick.

Place your microphone at a nice location. Now place your monitors/speakers such that they point towards the mic, forming an equilateral triangle. (In other words, the distance from the mic to monitor A is equal to the distance from the mic to monitor B.) Now flip the phase (using a plugin, or a switch on the speaker if you have it) on one of the monitors.

What happens now? Well, the sound from monitor A will reach the mic at exactly the same time as the sound from monitor B. But because one of them is phase-inverted, they cancel each other out, causing NO sound to arrive at the microphone!

If you do this, you can simply play the cue mix through your speakers and sing over that! You will hear it, but the mic will not (or not loud enough to be a problem).

A very natural way to sing. I really like it, although it takes some effort to get the phase cancellation working well enough.

Mind you! The “tweeter” in a monitor should be on your ear-level! Otherwise, you still hear pitches the wrong way. (The tweeter is the small bit at the top of a monitor that produces the higher frequencies.)

Conclusion

These are all the things I tried and what I think of them.

In the end, I learnt a lot from these experiments, and settled on a combination of everything:

I’m still learning, still practicing with headphones and recording, but these insights already made me stop hating every vocal recording (of myself) and created some vocals that I’m actually proud of :p

And after that, it’s just a matter of doing your vocal exercises and training your voice to be more consistent and powerful. (And, of course, writing good songs and all that.)

No matter what I try, I’ve never been able to make autotuned vocals sound human and lively. At least, not when they are 1/4 note (or more) off! Only very minor adjustments, if a note is really bugging me, work great. Otherwise, I suggest you do everything you can to get the perfect take right from the start.

You’ll know when you get it right, when you can play the vocal on top of its accompaniment (such as a guitar, or piano) and it sounds right. It doesn’t sound dull, or lifeless, or wrong, or clashing. The vocals match the pitches of the instruments, and that should work beautifully.

That’s all for today!

Until the next time, El Troubadour (Tiamo)

Update! Some people also suggest sticking a finger in your open ear (that doesn’t have a can), or a foam earplug, or something similar. This would, in theory, cause us to hear ourselves and the pitches better as well. But I haven’t tried this yet!