by Pablo Bellinghausen –
As seen in our previous instalment, recording voice is perfectly achievable in a home studio setting as long as the correct microphone is used and the room’s acoustics are taken into account. There are however a few other pitfalls that should be avoided if one wishes to attain professional results.
A good sound engineer can take home-recorded vocal takes and make them sound like a finished product by applying the necessary effects during mixing, but only if the recordings are clean, consistent, free of hum, noise, distortions and clicks, and either dry or made in a good-sounding acoustic space.
A basic recording setup with pop filter, music stand, monitoring headphones, and fancy “digital sheet music”
Although we won’t be looking at the vocal effects themselves this time around, it is important to know what the eventual processing chain is going to be, which will depend on the type of vocal. For example, an audiobook recording would usually aim for a clear, dry sound with no noticeable effects, a radio voice-over should have a smooth treble, a flattering bass and some strong compression to achieve that “larger than life” broadcast sound, and any popular music singing would go through several stages including equalisation, compression (sometimes several in a row!), pitch shifting or tuning, distortion, reverberation, delay and modulation.
With that in mind, the basics of a good take are pretty much the same independently of the recording style, and the first thing to do is get the setup right, which means giving some thought to the accessories around the microphone.
Mic stands are often neglected but using the wrong one can make a session uncomfortable, and depending on the setup, lead to expensive consequences; stands that droop, tip over or drop a mic from several feet high account for many broken condensers.
Since singers and voice actors tend to achieve better results while standing up, normal floor stands will do just fine for stage dynamic handhelds and lighter condensers; heavier valve models will however require either a stand with no boom, or a heavy-duty one with counter-weights.
The RØDE Procaster mic and its PSA1 multipoise boom arm are a good combo for spoken-word recordings
Spoken word, whether in podcasting, announcements or voice-overs, is very often recorded while the speaker is sitting down; in that case a good table stand (ideally a high-quality multi-poise model) is a better choice. It is a good idea to make sure the chair doesn’t squeak and that the clothes don’t rustle against the backrest, and to consider a thick table runner or blanket to tame reflections from the table itself.
The JZ|PF pop filter is a space-age variation of a classic accessory, but the results speak for themselves
A pop filter isn’t always required for models that have some built-in protection inside the grill, such as most broadcast and stage dynamics, but it is almost always a good idea – and pretty much a requirement when using studio condensers and ribbons (particularly at close distances). A pop filter can be placed anything from one to seven inches from the mic; a smaller distance will allow the performer to get closer and therefore get a drier sound, but will make it less effective.
The basic ones will be made of nylon fabric; steel models can be more easily cleaned, but be careful of buying cheaper ones since they can sometimes resonate with the vocal, adding an unpleasant “whine” to the sound.
Foam windshields are meant for constant low-speed wind and not the high-speed gusts our mouths emit when pronouncing ‘P’ or ‘T’ sounds, so they will tend to be ineffective, as well as affecting the sound negatively.
Setting up the Space
If recording with a computer, steps should be taken to ensure there isn’t any intrusive fan noise; although most well-designed laptops should be almost inaudible, it’s always best to put them as far away from the microphone, ideally in the direction where it is least sensitive (directly behind a cardioid model, for example), or, in cases where the computer is particularly loud, either behind the desk, covered with blankets, or in another room altogether.
Although a small amount of fan noise will be unnoticeable in a busy pop or rock mix, it can easily be distracting in spoken word, or on an exposed vocal in sparser music styles. The problem can to a certain extent be alleviated through well-applied noise gating, but a very noisy recording is almost impossible to salvage even with advanced noise reduction effects, so a little care during this stage can be critical to the end result.
The Audio Technica ATH-M50x has rotatable cups which make it easy for singers to use only one of them
Monitoring isn’t strictly necessary in spoken word, but it is critical when recording any type of singing. The recording engineer should spend a bit of time getting the headphone mix right for the singer, even at the expense of his own if there is only one headphone mix available, adding reverb and changing levels if required. This means the singer should use comfortable headphones that have a good midrange and don’t bleed (since this will be picked up by the microphone). Sometimes singers prefer using only one of the headphone cups in order to hear their voice inside the room; care needs to be taken with the bleed from the other side, and ideally it should be turned off completely for that reason.
A very common problem that can completely ruin a whole session if one isn’t careful is distortion.
Everything in the audio world, from microphones to interfaces to computer files, has an ideal level range in which it sounds best, and optimising the level of a signal throughout the sound chain, called gain-staging, is a very important part of professional recording. In most cases microphones can take louder signals than most performers (with some notable exceptions) are able to produce; the problem in home studio recordings usually arises at the input stage on the audio interface.
Most people are used to seeing a compressed, high-level signal on their recording software and will instinctively put the gain up as high as possible. This is usually a bad idea; the difference in level between a soft vocal at one foot away from the microphone and a scream a few inches away from it can be upwards of 30 dB, and setting the gain too high is almost certainly going to cause horrible clipping distortion during louder sections.
Singers can have a large dynamic range; setting up the gain too high can cause clipping in loud sections
The best way to check the correct gain is usually to ask the performer to speak or sing as loudly as they are realistically going to get, as well as getting a bit closer than they would for the actual take, and set the gain with that in mind. If the sound wave on the recording software looks too low but the playback sounds fine, then don’t worry; careful post-processing including compression and manual “fader riding” from the sound engineer will take care of raising the volume; analogue-to-digital converters inside good audio interfaces nowadays have a 24-bit resolution, and very low self-noise, so in the large majority of situations the sound of the room will be louder than any internal hiss if set at a conservative level. If hiss is a persistent problem in your recordings and it’s not coming from inside the recording room, then either your microphone or audio interface might need replacing.
Once all of the above steps have been taken it will be hard to get a poor-sounding vocal, but there are still a couple of considerations to make, mostly with regards to the position of the performer in relation to the microphone; it should be kept as consistent as possible, since microphones will sound different depending on the distance and the angle. Singers will naturally move around a little bit, and this can often help getting an expressive performance, but the more they do, the more unmanageable the differences in sound will be, particularly when trying to match several different takes of the same part.
It is important to pay attention to breath and lip noises, which can get amplified in close-up recording and sound distractingly loud after processing. Breath noises are a part of a natural-sounding recording and so they should not be completely removed, but the performer should keep them in check and try to minimise them whenever possible.
An experienced singer will often have a well-honed microphone technique, which will mean moving around the microphone to help consistency instead of fighting it, by moving closer in quiet, intimate moments and further away during louder ones. However, when in doubt, and particularly in spoken word recordings, staying in one place will avoid distracting changes in sound and will make the mixing engineer’s life a lot easier.