Getting the Best Vocal Recording

Almost all pop music features someone singing. Getting the best vocal performance is crucial, since most listeners respond to the words of a song and need to be able to hear them.

In this episode, I first discuss the human factors involved in getting a great performance from singers, mostly from a producer’s viewpoint.

And then I focus on the technical requirements of capturing that performance using microphones, mic preamps, processing, effects, and mixing. Often there are problems to be fixed in a vocal track, and I review some of the techniques from salvaging a performance that is less than perfect.

email: dwfearn@dwfearn.com
www.youtube.com/c/DWFearn
https://dwfearn.com/

Transcript

93           Getting the Best Vocal Recording                                              May 28, 2024

 I’m Doug Fearn and this is My Take on Music Recording

 Humans have been talking for at least 200,000 years, and probably singing just as long. The human voice is the primary way we communicate with each other.

So, it is not surprising that the majority of music features someone singing. And even music that is entirely instrumental often has a “singing” component.

Louis Armstrong said that his coronet playing was simply him singing, using his brass instrument. Indeed, we often respond positively to instruments that mimic the sound of a voice.

Clearly, our musical heritage is based on singing. It’s no surprise that almost all popular music relies on a singer to convey the message of the songwriter.

Since the singing voice is such an integral part of most of the music we record, we should strive to get the best possible version of the vocal.

And I believe that doing that job well also includes getting the best performance out of the singer. As an engineer, your input on that might be minimal if there is a strong producer in charge of the session. Or if the singer has strong ideas about their performance. But if you are acting as the producer of the song, you can influence the success of a vocal performance.

I’ll get into microphones and such later, but first I want to talk about the art of getting the best performance from a vocalist.

I should mention that many of these things apply to any musician, playing any instrument.

The studio can be an intimidating place for many performers. Even if they have had years of recording experience, they still perform under the weight of knowing that what they do that day may live on forever. They want it right, as do we all.

To help them, we have to provide the atmosphere where they can do their best.

I have found that even top singers respond well to your sincere approach to recording them. I make it a point to get to know them, at least a little, before we hit record. I do that in the studio, or other comfortable place. It doesn’t take long. Just look at it as two human beings making a connection.

You need to let them know that you are there to help them get the best performance possible. You’re on the same side, with the same goals. You don’t actually say those words. Just give that reassurance to the vocalist and they will likely respond well. The best engineers are always excellent at dealing with other humans.

Next, you should make the studio a welcoming place. Some singers like an audience in the control room. Others want no one extraneous hanging around.

It’s a bad situation when extraneous people are hanging around and not involved with the vocal recording. You might have to remind them to stop talking and help the singer by eliminating distractions. Often, those people will leave the control room or studio at that point.

The studio should be set up to help the singer relax. Ask what they need. Sometimes they do not know, so it may be up to you to determine what will make them comfortable.

Do they need a music stand? Where would they like it placed?

Do they want to face the control room window? Or face away? Do they want lights dimmed? 

I have worked with singers who want near-total darkness, and others who want a bright and exciting environment.

Make sure the temperature in the room is comfortable, and change it if necessary.

Be sure they have what they need. It could be a bottle of water. It could be a cup of tea. Try to accommodate them.

For most singers, the vocal performance is a very personal thing. They are putting themselves out there for the world to hear. If they wrote the song, their involvement with the music is even deeper. The lyrics mean a lot to them.

No matter how serious the song is, there should always be an atmosphere of fun. It doesn’t have to laugh-out-loud fun. The experience should be enjoyable for them. You rarely get a good performance from stressed-out people.

To help them, you may need to help them relax. And that applies to high-energy songs, too. They don’t have to be half-asleep, but they should be comfortable so that that mechanics of getting a good take are fun, not stressful.

Many singers start out tense. If you are the producer and know the artist well, it may only take a few words to remind them to loosen up so their voice does not sound pinched.

With singers you have worked with previously, you should be able to set the levels and any processing ahead of time. It may take a small tweak here and there, but it should sound good from the start. Keep in mind that singers usually end up being about 6 to 10db louder when they reach their warmed-up level. You can set the initial levels accordingly.

The headphone mix and volume can make a difference. I have talked about that in a couple of previous episodes.

For any performer in the studio, they will do their best when what they hear back sounds great. If you have done everything right, it will sound great to them right from the start. That will inspire them and you’ll get a better performance.

If it sounds lousy, they will be discouraged. Don’t let that happen.

I always record the first run-through. You are almost certain to discard it, but it can be helpful for the singer to listen to what they did. Most will hear right away any adjustments they need to make in their singing.

If getting it right has become a struggle, listening back to the first version may help clarify some things for the vocalist. You might even end up using that performance, in its entirety, or parts of it, in the final mix.

There is something about the “throwaway” quality of a first run-through that may be the best take you are going to get.

But usually, it takes a few times through to get the best performance. Know when to stop trying. This is a decision made by the singer, the producer, and sometimes the engineer. Every performer will reach a peak point and then the quality will start to go downhill. Maybe a break will help, but usually I find that that does more harm than good. Singers often have to keep multiple things in mind as they sing, and a break can ruin that concentration. Occasionally, you might have to come back another day, if that is an option.

One problem is focusing on the details of the performance and not the overall impact. It is easy to get so obsessed with getting every nuance correct that the main purpose of the vocal is lost.

Sure, you can make a composite version out of many takes, picking a line from this take and a word from that one. Digital recording makes that simple. However, I always strive to get one great take from beginning to end. There may be a few imperfections in that performance, and you have to decide if those are fixable or if they are minor enough to ignore. This may be a collaborative decision of all involved. But the engineer has to know what is possible to correct and what is not. For example, if the singer changes the phrasing, or the words or melody slightly, it might not work with a previous or subsequent take.

Some singer-songwriters will change the words from take to take. If you notice that, ask them if that was what they intended. Often the singer is unaware that they changed something. Usually this is only a concern if it affects the storyline. But changes might make a composite vocal more difficult or impossible.

We have the tools available to salvage a take that has the right spirit and feel, but suffers from some problems. This has made perfection the goal for most people. I’m not sure that is the best approach. Sure, sometimes there will be just one note, or a few, that are sharp or flat enough to be a distraction. Those can be fixed, either through pitch-correction software or by finding another take that has that note more in-tune.

I am always more interested in the emotional impact of a vocal than I am about its perfection. It’s great when you can have both, but I will always go with a bit of imperfection if that is the best take to connect with the listener.

Think about great vocal performances from back in the day when everything was recorded at once. There was no option to punch-in, pitch correct, or comp a vocal from multiple takes. They might have recorded the song a few times before they got a great performance out of everyone, but that was as far as they could go.

Those performances were imperfect in many ways. Singers were off pitch, or flubbed a word. Their phrasing might be inconsistent. That did not stop the song from being a hit, and a classic that is still valued today. The performance affected the listener in a way that was meaningful to them, and that is the most important thing.

Give the singer positive feedback when necessary. Telling them they sound great is always helpful. Just make sure it is true.

Should you give them guidance if you think you know what would improve the performance? Maybe. It depends on the people involved and your level of experience.

Some singers are grateful for suggestions. It will just annoy others. You have to determine which applies to the person you are working with.

Some producers will be very unhappy if you suggest something directly to the singer. They may have different goals in mind than you do. You might not even understand what they are going for. But if you hear something you think could be improved, tell the producer privately. They may welcome your suggestion. They may even rely on you to provide that input. Their focus is elsewhere. In any event, let the producer tell the artist.

With a producer you have worked with a lot, they may be OK with you making those suggestions directly to the artist. It depends on the people involved. Whatever you do, don’t contradict the producer, or give the singer suggestions that conflict with information that the artist and producer have already discussed.

This is a delicate area, and sometimes you are just going to have to cringe and live through a situation you feel strongly is a bad take, or bad approach, or bad advice.

It’s much simpler if you are the producer. You should be the one making those decisions. But listen to your engineer, too. They can be a valuable source of experience.

And just a word about criticism in your role as producer. If the vocalist seems to be having a bad day, you might have to keep quiet and just try to get the best take you can. But if you recognize a broad problem, it can be useful to the singer to discuss it with them.

But don’t do that during the session. There may be exceptions, but during the session is not the time to try to fix big problems. As always, it depends on the people involved, and it is your job to understand the personalities.

My rule is to avoid bringing that up during the session. Or even at the end of the session. Wait at least a day, and preferably longer, and then diplomatically bring up the topic.

It should always be collaborative, not one-sided. You usually cannot dictate a solution and expect cooperation. Explain that you have been thinking about their vocal and you think you might have some ideas about how to improve their performance in the studio. If the artist has trust in you, they should be glad to hear your thoughts. Every creative person should want to learn new things to improve their art. This is entirely dependent on the personalities involved. Don’t make the situation worse if the collaborative spirit is not there.

And during the session, if you have a minor tweak to help the singer, think about what you are going to say and phrase it in a way that builds trust with the artist. Perhaps it is the pronunciation of a word, a change in phrasing, or the emphasis on a particular word. I think those types of things are appropriate for a producer to bring up during the session.

But another rule I have is to never tell a singer to change more than one thing at a time. During the take, you may have made a list of nuances you think need addressing. Pick the most important one and work that out before going to the next one.

Some singers can take two or even three suggestions at once and make the changes easily. But most performers can’t do that comfortably. You certainly do not want to compromise the performance by making the singer think about the mechanics too much. As always, this depends on the people involved. You need to develop the skills to know how to shift the performance in a way that is not stressful to the singer.

Now let’s talk about the mechanics of recording vocals. An early decision is what microphone to use.

If you have worked with the singer before and found the mic that works best with them, you can simply set that up and start recording. However, if the song is very different from what you have recorded with them in the past, you might want to consider whether your “standard” mic still makes sense in the new context.

With a new voice, I find it helpful to listen to them in the studio a bit, if they are willing to run down the song at that point. Or maybe they have a demo of the song you can study ahead of time. Experienced engineers will quickly get an idea of what mic is going to work best.

Do they sing loudly or softly? That might influence your choice. Do they have a wide dynamic range? That could be intentional or not, but you will need to accommodate that.

Is their voice harsh? Sibilant? Distorted? Breathy? Loud?

Is their voice out of balance with too much or too little bass? Midrange? Highs? Those characteristics will influence your mic choice.

Do they have good mic technique? You won’t know this until you set up a mic, so you might have to wait to see what happens and make a change if necessary.

If you have a wide collection of mics, you may have many options to try. If you have only a couple of “vocal” mics, your options are more limited.

You might want to set up a few candidate mics and have the singer sing a few lines of the song into each, with the track, and see what seems best. I would be careful with this, however. You don’t want the mechanics of the studio to distract the artist.

On the other hand, if you think you may need to try different mics, have them set up and ready to go ahead of time. You don’t want to interrupt the artistic flow by taking ten minutes to set up another mic. Most singers are fine with a bit of experimentation, but don’t overdo it.

Just because a certain mic has a long reputation for being a great vocal mic does not mean it is going to sound best with the singer currently in your studio.

I’ll give you an example. When I first started working with Corrie Lynn Green, she had virtually no studio experience. In fact, she had never sung into a microphone in any situation.

I knew her voice from the demos she made with her phone, and I had a fairly good idea of what I might use to record her. My first thought was an AEA R44. But would that be a good mic for a beginner with no experience? I had some reservations, but I was delighted to find it sounded great on her voice, and with a couple of suggestions, her mic technique was excellent right from the start.

I was totally happy with the sound, but was it the best choice? I wanted to try some other mics, if for no other reason than to confirm my initial choice.

I set up a U47. I almost always use my variable pattern condenser mics in the bi-directional setting, and that’s what I tried on Corrie.

It sounded good. But not as good as the R44.

Next, I tried a M49, which I like on many voices. On bi-directional, of course. It, too, sounded good.

Either the U47 or the M49 would have been good matches for her voice. But the R44 brought something special that complemented her voice perfectly.

She heard it, too, when I played back the samples of the three mics. There was no question in her mind which one she liked.

So that is what I used for that session. And for every session since. Along the way, I did try different mics when I thought they might add something for a particular song, but the only other mic I ended up using on a song was a C12 on one section of one song where I needed a more aggressive sound to contrast with the other, quieter, parts of the song.

By the way, those three excellent condenser mics were all made by Flea in Slovakia.

Is there still another great mic out there that might be better still? Perhaps. But I like the intimacy of the 44 and its super-smooth sound. It is perfect for Corrie’s voice.

As much as I like the R44, it’s not the right choice for every singer. It comes close to being my universal vocal mic, but some of the people I record sound better on a different mic. In addition to the U47 and M49, over the past few years, I have used a Horch RM2J, a Shure SM7B, an AEA KU4, a Coles 4038, a Bock 251, and even a Sennheiser KMH416 short shotgun mic on singers. Each choice was based on the singer’s voice, the song, the accompaniment, and the need to be able to hear the words.

In the 1970s, I used a Neumann U87 on most vocals. It was the standard back then. The older standards, the U47 and the AKG C12 in its various incarnations were also popular. But I recall one session where a Neumann KM84 was the best choice. The KM84 is not thought of as a vocal mic, but sometimes something unusual turns out to be the best thing.

On every sound, I first try what seems to be the best choice, based on my previous experience. That’s usually fine. But if there is something not quite right about that mic, I try to analyze what it is that I don’t like about it, and then think about what else in my collection might address that problem.

It’s true that any decent mic will give you a viable vocal sound. Even live-sound mics. But if I can find something that works really well, why not use it?

The right mic choice means you will have to do little or no processing on the mic. On ribbon mics, you will probably need to roll off the low end a bit due to proximity effect. But beyond that, I can’t recall using any eq on a vocal track in decades. I am a firm believer in doing as little to the sound as possible. The result always sounds best that way.

You might have to use some eq on some voices with some mics, but a range of available mics usually makes that unnecessary. If you need to use a lot of eq, it is an indication that the mic is not the best for the singer.

The mic preamp will have a large effect on the sound, too. Not as much as different mics, but the right preamp could make the difference between a lackluster vocal sound and a great one. A good preamp will make all your mics sound better.

As someone who designs mic preamps and other outboard gear, I naturally prefer the sound of the products I designed. After all, I designed them to fit what I thought sounded best. But your choice may be different. Find the sound you like for that singer, on that mic. If it sounds right to you, it is right.

And what about compression? I find that a touch of compression when cutting the vocal often helps keep the dynamics under control. You don’t need to slam the compressor in most cases, unless you are looking for that sound. Just a couple of dB of compression on peaks, with a nice, civilized compressor helps the vocal stay above the instrumentation.

Most experienced singers have good mic technique in the studio. Someone new to recording may not, especially if they are used to working with a live-sound mic on stage. You may have to explain the difference.

I find that most singers sound best if they are at least a foot away from the mic. There are exceptions, but usually that is a good starting point. That distance will eliminate a lot of the problems, like popping the mic or excessive sibilance.

Remember the inverse square law. With a microphone, it states that when you double the distance to the mic, the level drops to one-quarter the level. And when you half the distance, the level goes up by a factor of four. That’s a lot of change, and it will be audible.

Actually, with directional mics, the level variation is not quite that severe, but it is still significant.

Maintaining a constant distance will make your job easier and will probably sound the best. But there are exceptions. A really intimate song may work better with the singer closer to the mic – assuming that they can do that without introducing other problems. And remember the inverse square law. The closer they are to the mic, the more critical it becomes for them to maintain a constant distance.

In some songs, there may be a need for the singer to really project on certain parts, to best convey the emotion of the song. In that case, experienced singers can back away slightly during the loud parts. I like that effect in some situations because the room sound will go up as they back off, which can enhance the impact. If you have a good-sounding room, that is.

All directional mics have proximity effect, which increases the bass content of the vocal as the singer gets closer to the mic. Ribbon mics, in particular, have this effect. And it varies from one mic model to another.

It might seem to be a problem, but it can also be a tool to be used. Experienced singers throughout the decades have used proximity effect to change the sound of their voice in appropriate places. It can be a compelling effect. But it’s tricky and not every singer can pull it off effectively.

Should you use a pop filter or pop screen? I do not, except maybe in rare situations. I don’t like the change in sonic quality that occurs. But sometimes you will have little choice. Some singers can pop a mic on vowels, their voice has so many plosives in it.

There have been occasions where I have used a pop screen simply to define the minimum distance the singer can get to the mic. In that case, use one with the largest open mesh. If this were a frequent problem, I would probably devise some sort of “distance inhibitor” that clearly prevented the singer from getting too close, but had no effect on the sound. A pop screen with just the hoop and no screen, for an example.

Another technique is to angle the mic, so it is still aimed directly at the singer’s mouth, but it is not directly in front. With most mics, there will be little or no change in the vocal sound. But it will reduce pops, sibilance, and mouth noises.

I use that technique with the R44 I use for recording this podcast.

It's usually best to record a vocal by itself, but sometimes you have no choice but to record the vocal and the instrument the singer plays – at the same time. This is always less than optimum, but it can be done effectively.

You might think that cardioid or super-cardioid mics would be best for this, but I have found that bi-directional mics provide the best isolation, when used properly. Remember that the null off the sides of a ribbon mic are theoretically infinitely deep. That null extends all around the mic, in the plane of the ribbon, so it applies to the top and bottom of the mic, as well as the sides.

Multi-pattern condenser mics have deep nulls when in the bi-directional, or figure-8, pattern. Ribbon mics are even better.

The trick is to put the instrument in the null of the vocal mic, and the vocal in the null of the instrument mic. That is easy to do with a singer playing an acoustic guitar or similar instrument. It’s a somewhat different problem with a piano, but the principle can still be used.

One great advantage of bi-directional mics is that the off-axis sounds are still well-balanced in frequency response. That is not true of most cardioid mics, which become nearly omnidirectional the farther off axis.

And remember the inverse-square law. Keeping the mics as close as practical will certainly help with the isolation.

Close-mic’ing may involve compromises you don’t want to make, so think this through and be ready for some experimentation.

So why is the isolation so important, anyway? Well, the differences in the arrival times of the two sound sources will result in a phase error. At its worst, you might partially null out the desired sound. You can try inverting the polarity of one of the mics with the phase switch. It may help. Or it might just be different. At least this is one modification you can exactly un-do later if necessary.

Usually, however, the bleed between the sources and mics will simply reduce the apparent presence of one or both sounds. That is almost always a bad thing. The vocal or instrument will sound distant, no matter how close the mics are. Improving the isolation will help.

If the singer’s voice and their instrument are in good balance in the room, then achieving good isolation should be easy. I have found that I can easily get 20dB of separation between the two sources. That’s not even close to recording them separately, but it can work well.

Another approach is to record both the vocal and the instrument with just one mic. You need a performer who can play in perfect balance at all times, and many singers can do that amazingly well. You can’t change the balance later, of course, so make sure it is right. That gets trickly if other instruments are going to be added.

What about recording the vocal in stereo, with or without a simultaneous instrument? It’s still a “single-mic” as far as balance is concerned.

Stereo vocals can have an amazing effect on the overall sound of the recording. Even though you might think of the voice as a point source, it isn’t, really. And the contribution of a nice stereo room sound might further enhance the sound.

But I rarely record vocals in stereo. That is odd, because I often record almost all the instruments on a song in stereo. But I find that the solid, mono, vocal just works better most of the time. And in stereo, the singer has to stay perfectly in the center of the stereo image for the entire song. You don’t want the vocal to wander from side to side. A little bit of that is OK, but I have found few singers who can perform at their best when staying perfectly on-axis to the stereo mic or pair of mics. Greater distance helps make this less critical.

In general, most songs have lyrics that need to be heard clearly to convey the message. It depends on the music. In some genres, you can bury the vocal and let the instrumentation carry the energy. Other styles need to have the words easily discernable.

Of course, that is a continuum, and you need to determine what the song you are working on really needs.

An arrangement that leaves no space for the vocal will require a higher vocal level in the mix. Or you might achieve at least some of the solution through mic choice and processing.

That presumes that everything else on the song is right. The instrumental arrangement might be the problem. Sometimes just leaving out certain instruments, perhaps only in certain places in the song, can help keep the vocal properly out front.

If the bulk of the energy from the instruments is right in the vocal frequency range, then it is more challenging.

If the dynamics of the backing instruments do not match the dynamics of the vocal, then that is a problem to be fixed.

It is always best to solve those kinds of problems with the arrangement.

But sometimes these problems arise because the musicians do not have a clear concept of the song. Or maybe inexperienced musicians have not yet learned how to play with a vocalist. Some will never learn, and you can only do so much to fix that.

I find it useful when using studio players to have the songwriter, or the producer, explain the song to the musicians. A lot of musicians do not listen to the words and don’t have a clear concept of the message the singer is trying to convey.

If you did all these things right to begin with, the mixing process should be easy. I have done songs where I found the right balance and just left the faders in one position all the way through the mix. The song mixed itself, so to speak. But only because all the ingredients for an effective final version were already baked-into the recording.

Often, we have problems to fix. This is especially true if someone else cut the tracks, or multiple people worked on various overdubs. You can’t beat a situation where someone has an overall vision for the song from its inception.

If there are words or phrases that are getting lost in the mix, you might be tempted to put more compression of the vocal. Sometimes that works, but most of the time, in my experience, the problem remains. Sometimes it is the attitude of the vocal that needs changing. Singers may have to learn how to vary their level of projection. Projection is not always the same as loudness.

With automation, you can micro-manage the vocal level and get every word to be audible. It can be time-consuming, but sometimes it is the only way. Make sure your manipulations are not audible. You do not want to create a new problem to deal with.

What I often do with a song that is challenging to get the vocal level just right, is to play the mix back at very low level. I then leave the control room and listen around the corner. It should be no louder than quiet background level. Can I understand all the words? Do I need to bring the vocal up? Or maybe the vocal seems too loud in that listening situation and I have to reduce its level.

Or maybe the instruments are getting lost when the words are easily understood. You may have to untangle all that in the mix.

Are there important words getting lost? Can I automate the level as needed to fix that?

Is there an instrument that is masking or conflicting with the vocal? What do I need to do to eliminate that problem?

The song might lend itself to some effects on the vocal. A delay can sometimes help. It can make the vocal seem subjectively louder. So can changing the amount or the character of the reverb. Or using no reverb at all. The song might suggest a more heavy-handed effect on the vocal. There’s no formula for this. Do what serves the song best.

If you take the same approach to all the instruments on the recording, you will find that additional processing may not be necessary. It might even detract from the impact.

Most of the music I record is acoustic-based. I do record other genres, like rock, jazz, punk, or R&B, and world music, and I enjoy doing all those sessions, too. But acoustic music is my favorite. And I find that all I need for the mix is a bit of buss compression and overall eq, plus perhaps some level automation, and I am done. The finished product sounds full, appropriately loud, and you can understand every word.

That can be applied to whatever you are recording, no matter what the genre. Good musicians are required, but when all that comes together, you can focus on getting the perfect vocal performance. And you won’t have to worry about how you are going to fix it later.

Thanks for listening and commenting. If you know someone who you think could benefit from this podcast, please let them know about it.

Send any comments or suggestions to dwfearn@dwfearn.com.

Thanks.

This is My Take on Music Recording. I’m Doug Fearn. See you next time.