Vocal recording sessions start long before your singer turns up to work. Through this tutorial, we'll look at all the factors that contribute to a successful vocal recording session, focusing on preparation as well as the technical aspects of a session. To stress, we'll only focus on recording vocals here; we'll look at vocal mixing techniques in an upcoming feature.

In terms of preparation, ensure that your recording setup is ready in as many ways as possible before your vocalist arrives. This means making sure that your microphone is plugged into your audio interface or recording channel and that this signal is, in turn, being sent on to your computer. If you're working with a condenser microphone, you'll need phantom power to be enabled. So an hour or so before your singer arrives, make sure everything is working well, as this will give you time to trouble-shoot any problems.

One of the issues you might discover when testing your own voice this way is latency. Latency is the lag between a computer receiving an input signal and then playing it back as an output signal. As computers have the capacity to buffer the sounds they play, this lag can be substantial. During mixing, if you're running lots of CPU-sapping plug-ins in a larger session, turning the buffer size up to maximum can be beneficial, as the increased thinking time for a computer can reduce CPU overload. But during a vocal session, singing is extremely difficult if a performance is complicated by an audio time delay. So while recording, keep buffer sizes low and, if your DAW has one, enable its low latency mode, too, which may even more significantly reduce latency issues. Again, you can test this before the singer arrives by monitoring short percussive sounds like finger clicks; if you can hear a lag before the click plays back, try to find more latency reduction in your system.

This also has implications for the backing track to which your vocalist will perform. If you have a busy session with lots of software instruments and effects, your computer will need a larger buffer size to generate the sounds required. In preparation for a vocal recording session, it's worth bouncing your backing track to a single stereo file or a small number of audio stems, complete with all effects. Importing these into a new session specifically for the recording will save CPU power, and once the session is complete you can import the vocals from it back into your host project. Usually, bouncing a single stereo file will suffice, but if you think your singer will want more flexibility, create a few stems covering drums, bass, harmony elements and so on. The volumes of these can then be balanced to the singer's taste.

Below these stems, create a number of audio tracks for the vocals you're going to record, all set up with the correct microphone input number. If this sounds like overkill, know that inadequate preparation is one of the main factors in derailing sessions. Your singer will want to feel as though he or she can focus exclusively on singing, and every time you say, "Hang on, let me just set up another audio track," or "Sorry, that channel was muted," the performance bubble is popped.

Having stems and audio recording tracks ready are one consideration, but make other preparations, too. Singing is thirsty work, so have a jug of water to hand so that you don't have to keep running off to refill a glass. On the subject of drinks, avoid choices that could impact on singing quality or concentration levels. Caffeinated drinks aren't a great idea, as the initial buzz these provide gives way to fatigue, and it's worth avoiding anything too carbonated for obvious reasons. If possible, encourage your vocalist to stick with water until the session is over.

The goal of the recording environment should be to produce as dry a recording as possible. In professional studios, acoustic treatment is used to ensure minimal reflections, and you can bring some techniques from these into your space, too. sE Electronics' Reflexion Filters are a useful starting point for helping to isolate a vocal. Equally, creating a "dead zone" in one part of your recording space, either with dedicated acoustic panels or soft furnishings, can make a huge difference. It might look unprofessional to build a tent around a microphone with duvets and blankets, but it'll make a big difference to the sound you record, particularly if your studio is full of reflective surfaces.

Microphone choice will also be critical. If you're intending to do a fair amount of vocal recording, investing in a quality mic is essential. The good news is that good quality is now available for a fraction of the cost of a decade ago. For studio recording, a large-diaphragm condenser microphone will serve you best. It should be held on a microphone stand, ideally with a shock-mount and fronted by a pop shield. Shock-mounts act as shock absorbers to prevent reflections from carrying to the microphone; pop shields break up plosive p and b sounds to prevent the mic from receiving quick bursts of air that turn into explosive bangs. Even if you decide to record with a dynamic microphone (traditionally used for live recording rather than used during studio sessions), mount the microphone so the vocalist doesn't have to hold it.

It's also worth considering that some producers like to record vocals with a little compression. This will only be possible if you have a hardware compressor either built into the channel strip through which you're recording, or if you're working through an interface that features a software compressor whose output can be recorded as well as monitored. If you're recording a singer with a wide volume range this can be an advantage, but be warned that it's extremely difficult to remove compression from a recorded sound once it's printed. So err on the side of caution and either record without compression at all or, if you do use some, go gently. You can always add more when mixing later. Otherwise, don't record other effects as part of the vocal recording session if you can help it, steering clear of spatial effects like reverb and delay completely.

This leads on to an important consideration: vocalists often feel safer if they can hear reverb while recording. Fortunately, most DAWs allow you to set up an auxiliary reverb to which you can send your vocal recording channel while tracking, without this effect being captured as part of the session. To do this simply route the input channel in your DAW to a reverb and check with the vocalist how much of this effect they want to hear while recording. Both the send level from the recording channel and the auxiliary effect return fader will allow you to control reverb level. Effects used this way—heard while recording but not captured as part of the recording process—are called monitor effects. Using the native effects that ship with your DAW rather than (usually) more CPU-taxing third-party plug-ins should prevent latency resurfacing.

All of the above can be considered before your vocalist arrives for the session, and addressing all of these points should ensure that you're not faced with too many surprises once recording starts. To ensure that only the vocal is captured (and not the backing track to which the singer is performing), make sure your singer is wearing headphones and that your studio monitors are switched off. To hear what she's singing, you'll need headphones as well, so either a headphone splitter or an audio interface with more than one headphone output will be useful. When your singer arrives and you're running the track through, simply arm the recording channel initially, so that the microphone is heard live through that channel. Press play rather than record, and the singer can practice against the backing track.

When switching from practice mode into the beginning of the recording process, start by capturing a test take. So that the pressure is off, stress to your singer that you're recording this for your sake only to check levels. As this vocal is recorded, listen carefully and watch the levels going into the computer. With digital recording you're looking to capture the best signal-to-noise ratio you can, which equates to the loudest performance you can capture without the vocal tipping over the upper volume threshold into distortion. If the peaks in a vocalist's performance are reaching -3 or -4 dB that's an optimum level, but do bear in mind that through the first few minutes of a session, as the singer's voice warms up, he or she is likely to get louder, so keep the input gain dial at hand. There's one critical point here: to adjust vocal recording levels, the mixer inside your DAW won't help you. This controls the volume of the vocal on the way out of your computer after it has been recorded and doesn't in any way adjust the input volume. If the vocal starts looking a little loud, turn down the input gain on your interface rather than the volume fader in your DAW.

Some DAWs only allow you to record one vocal performance per channel, whereas others allow you to build multiple takes on a single track. If you can only record one file per channel, make sure you set up multiple audio tracks in advance so that you don't have to make a new track for each recording. However, if your DAW creates take folders, you'll only need a single track, as each new one will be stacked on top of the previous one. The advantage of this is that you'll be able to choose your favourite parts from each performance to build up a "comped" take once recording is complete.

Backing track: For the vocal recording session for this tutorial, a basic backing track was prepared that features a good balance between musical simplicity and vibe. There's not so much in the track that the singer will feel crowded out, but equally what's there offers enough to capture the singer's imagination.

Melody guide: As well as the synth and kick drum parts comprising the backing track, the vocal melody line was also played into Logic Pro X as a pitch reference. Sometimes it's helpful for a singer to have this safety net while recording. If you're comfortable singing, it is even more effective to capture a rough vocal guide instead, which has the significant advantage of pairing the melody with the lyrics. If not, playing the melody line in on a piano or synth will certainly help.

The session started with printed lyrics waiting for our singer. As she hadn't previously heard the track, we sang the song through together a few times until she was comfortable with the melody and then, to set up levels and check that she knew the song from start to finish, we recorded a first take. Capturing one of these is always a useful transition from the practice of running a song through to switching into a recording mentality. While it's unlikely you'll use this take, it serves as a check of recording level and allows the singer to request any changes to levels, reverb amount and overall headphone balance.

Having recorded this first pass, we agreed that breaking the song down into three sections would help focus on each one individually. We started with Verse 1, which we wanted to sound breathy and quite fragile. For this, a new audio track was used, and several takes were recorded. Between each one, if required, some constructive feedback was provided, either suggesting how a particular word might be approached or how long the tails on the end of each phrase might last. You don't have to talk between each take, and quite often a singer will want to get into his or her stride by recording several takes in a row. On the flip side, setting a loop around a phrase is rarely a good idea, as a singer might not understand why she is being asked to record the same part over and over again. So press stop between each phase and, unless longer feedback is required, offer a couple of words of encouragement before recording the next pass.

Verse vocal: Pretty soon, if you're recording to a take folder, you'll end up with a series of performances on the same track, with the most recent pass at the top. In my forthcoming feature on vocal mixing we'll discuss the best parts from each take, so this clip features the complete final performed phrase in raw form, free from any additional processing or take selection.

When you're confident that, across your takes, you have each line captured, move on to the next stage of the song. Make this a positive transition, resisting all temptation to say, "That one's OK and will sound fine once I've tuned it," or words to that effect. Choose your words carefully, as the quality of your track will depend on your singer's confidence.

Lead vocal complete: Having completed the verse vocal, the singer recorded the two halves of the chorus section onto separate tracks, having agreed that the first half needed to be a step up in intensity from the verse, while the second half needed to build even more. As the dynamic change from the verse to the second half of the chorus was substantial, the input gain was adjusted through these recordings to ensure levels were kept in check.

Floating chorus track: As producers, we tend to have a pretty clear idea of where a track is going—from our backing track, it's fairly apparent that the main drop of the track is going to happen right after the vocal ends, for instance. However, in terms of building expectation and mix density, think beyond a lead vocal to consider how other vocal performances could also help build drama. In the chorus section of this track, the vocalist and I agreed that a softer vocal performance could float gently under the main lead vocal to add a second layer. In this clip, you can hear this performance without the lead.

Floating chorus track, in stereo: It's particularly effective for gentle parts like this to be recorded twice, or "tracked," so that one performance can be panned to the left and the other can be panned right. This produces a wide stereo image with the lead vocal remaining panned centrally between them.

Whisper: Another effective technique is to ask your singer to whisper the lead vocal in time with the sung equivalent. This adds a mystery and fragility to the performance but it also helps with articulation, creating emphasis to help words cut through the mix. The final part we recorded was a whispered take of the whole song. Listen carefully and compare this take to the track above for a before-and-after comparison.

Monitor effects: It's always interesting to hear vocals in their raw, unprocessed state like this—it's almost uncomfortable hearing performances without effects to blend the vocals to the track. This will be true for your singer as much as it will for you, so before the session is over, it can help to spend five minutes reintroducing some quick effects to make your track sound a little more like the record it's going to become. This will help reassure the singer that he or she has done a great job, which builds confidence for the next time you work together. Here's the whole of the track again, with the session's monitor reverb reintroduced on the vocal tracks, as well as a subtle 1/4-note delay.

This leaves us in great shape to contemplate a mix of these vocals, which is where we'll pick up in the next feature. In the meantime, remember that a successful vocal recording session is dependent not only on your ability to run your computer's audio recording capabilities as second nature, but also on your communication skills. Any seasoned engineer will tell stories about nightmare recording sessions, and often these stem from one poorly worded sentence that caused offence, from which recovery proved impossible. Get a session right, and your vocal takes will provide a great platform for the rest of your production and mix.