Posted: 13 August 2004 at 4:34pm | IP Logged

Vocals – the mp3u guide to recording vocals by Fluid of Fluid Productions

Every musician aspires to that 'Pro sounding' quality in their final mix. This is significantly harder when you’re using your own home computer to try and achieve everything. You might get away with crystal clear recordings of all elements transferred to your computer digitally but you will not have any such luck with vocals. Vocals provide a difficult proposition for the bedroom producer. Firstly, if your computer is not powerful enough you will suffer latency issues and so forever be trying to rectify the vocal channel timing. What can you do? The higher the quality of your recording the worse the latency seems to get as you require a larger buffer for the recordings and samples. You can reduce the buffer size but this can considerably alter real-time performance and other important mixdown factors. The difference between a 24bit and 16bit recording can be significant and so may provide some relief. Now I’m not going to address this issue in this article, there are too many variables that can affect the latency problem (soundcard, memory, processor, etc…) but I am going to look at how we can simply address the quality of our vocal production regardless of set-up. For some of you, this may mean a reduction in latency because I will be talking of recordings in the standard PCM, 16bit 44,100Hz and many will have soundcards that default to 24bit 48KHz recordings. Consequently you can reduce the quality of recording and not have any formal loss in your final vocal. So let’s start achieving our pro sounding vocal.

For this tutorial you need a microphone, sound editing software (I'm using 'Cool Edit', its cheap and easy to use and can produce 'pro' sounding work although I prefer Wavelab and Soundforge). And also a sequencer – I'm using Steinberg Cubase, although Logic, Digidesign Protools and any other sequencer will do just as well…and I can hear some of you already saying ‘my sequencer will do it better!’.

Firstly to get a pro sound, you have to work at the source, the actual recording.

1: Make sure your microphone is sufficiently away from the computer. Most of us do not have a studio in our bedrooms so the hum of the computer will spoil the vocal takes we record unless of course you have invested in substantial component sound mufflers. Many of us haven’t and for us the following is useful.

If there's a slight drone then we have no problem, we can cut it out using the editing software, but software cannot cut out loud noise without damaging the vocal take. You could try recording outside the room with the door shut so that it also doesn't pick up anybody else who's in the room. Obviously this requires either very long leads (unless you have a very small room) or the placement of your computer in a strategically suitable place. That should do the trick. If you have a basic 'live' style microphone, then this step doesn't matter.

2: Make sure the microphone is on a stand. If the performer is holding the microphone there is no doubt going to be that slightly uneven feel and movement in the sound quality. A static microphone will overcome this problem. Remember, avoiding a problem is easier than dealing with it once it occurs.

3: Rap or Sing at least 6 inches away from the microphone, or get your vocalist to do so. That has proved to be the perfect distance in most cases. It can depend on the performance and the type of vocal being sung or rapped. The standard six inch gap will provide a nice clear vocal but you should remember that certain tones or notes will be better if you (or the vocalist) pull slightly away from the microphone. Those VERY big notes that divas typically hit are the type of note where it is best to pull away slightly. This will not diminish what is being sung and will avoid the problems that come with a sudden influx of big noise.

4: A ‘popper stopper’ is an extremely useful device to stop any unwanted tuts, esses or pops when the vocal is being performed. You can buy one. You can also easily make one, by wrapping some tights around the microphone. I used to get one of those metal hangers and make a wire circle around which I used to place quite heavy denier tights. My parter was never very happy about this – finding her clothes strewn on the floor and her latest tights cut to pieces but I found it useful.

Now that the basics are sorted lets start recording. I'm presuming you already know how to record with a microphone using your existing software so we’ll jump straight into a few pointers for recording and the all important post-recording process.

1: Open up your song and select an audio track.
2: Always start the track, two phrases or bars before the performer is meant to sing, so they have a little time to get ready. End the track two phrases or bars after the performer’s finishing cue so they can add any 'ad-lib'. You can always delete this final ad-lib if you don’t like it.
3: Record the vocals a minimum two times. The second time is usually better then the first but if you find that more is required then just keep going until the vocals are as good as you want them.

With this now done we can start getting that 'Pro' sound! At this point it’s all down to your sound editor.
Open up your recording in your sound editor!

When you first open a vocal audio file in Coo Edit you’ll generally see the same thing as everyone else…and this occurs because no human performer is like an automaton. The volume will be all over the place! Look at the waveform, it’ll go from really loud, to really quiet! The first rule is that we need the vocals to sound about the same volume throughout the rap/singing otherwise the listener will have trouble hearing some words…and unless this is planned you already have a disaster on your hands. So what can you do? Well a general rule that I learned from one engineer is called the ‘THREE C’s’, that is compress, compress, compress!

What’s a compressor and why should you use it? Well simply a compressor takes the audio and makes the loud parts really loud at the same volume and the quiet parts really quiet (at the same volume). All the vocals stay really loud, whilst all the breath, hum, and background noise stays really quiet. You should, as a rule of thumb, do this to every vocal you record. Open up your compressor in your editor, In Cool Edit it’s in the: Transform/Amplitude/Dynamics Processes. The great thing about cool edit is that it has presets! So we can simply go to 'Vocal Compressor 2' then press ok. For those who don't have this option, you should compress to 5.41 (threshold -24 db), expand 1.76 and compress again to 1.69 (threshold -67.4 db).

The final result gives a vocal waveform that is more-or-less even. It might seem a bit loud, so just quieten your audio (e.g. Transform/amplitude/amplify). Now the compressor also makes any unwanted background noise louder (and more apparent) so just highlight that sound, then silence it (e.g. in Cool Edit go to, Transform/silence) DO NOT DELETE or else this ruins the alignment of the vocals to the music. This single post-recording step will transform your vocal and clean it substantially, but we can – and need to – do more.

A popular method these days is to add some reverb! Reverb is a great way to add some acoustic flavour to the tracks you produce. Try and add it in subtle amounts or else the vocals will sound out of place. I’ve found a slight and ‘warm’ ambience is sufficient to lift the ‘flatness’ of the initial vocal recording. Since I’m focusing on cool edit let’s use a cool edit preset. Highlight the vocals and then go to 'Transform/Delay effects/Reverb' option. Add a 'Vocal Natural Reverb' and press OK. Listen to the wave, there's a slight reverb to it. If that's too mellow and understated try a louder one, or even add a slight 'echo'. Experimentation is fundamental to your development of a vocal style and tone that suits what you are producing. You should now have a clean vocal that will embed and enhance your mixdown much more appropriately. This is, of course, just a start. One of my methods would involve numerous other tricks and I shall point some of these out below.

Tricks!

1. Add an echo to some words, to give them some depth. Think of things such as echo as grammar tools in relation to musical dialogue. The punctuation that an echo can create can often be more dramatic and convincing than additional words explaining. If you viewed you lyrics as poetry and the performance is simply a reading FX like echo are the stresses and calms of your meter.
2. Ask the vocalist to back himself up, by recording again, but keep the other recording on another track. ‘Doubling-up’ is a standard procedure. It will allow the vocal to be thickened but also allows some other things. I personally copy the double track and then apply echo (as well as other effects) to small vocal snippets and then silence the rest of the vocal. I end up with a variety of vocal tracks for the mixdown that can total in excess of eight tracks based on the same lyrics. With harmonies and backing this total can often triple.
3. You could try adding some EQ. EQ can offer some great effects like the classic ‘phone' effect. It can also remove some of the unwanted bass contained in the vocals or even increase it…depending on what you’re trying to achieve.

Experiment! The technique of reversing a snippet, adding reverb and then reversing back had to be tried by someone in the first place. If you don’t try you will not learn and you will not develop your style and sound. I find it useful to listen to what other artists are doing and then try and break it down into a process. I take the ideas from this analysis away and try and emulate what I’ve heard. Sometimes it works and sometimes it doesn’t BUT I always learn something new about processing audio…and each new idea just adds to my armoury.

I hope the tutorial proves useful for some of you. Undoubtedly some will stick with the method they have and others will try what I suggest. Some will hate the results and others will love them. The truth is that those who haven’t developed their own recording style for vocals will learn something regardless of what I’ve said here because they will be practically doing something. I hope one of you comes back with some great new idea to add to vocal recording and processing that can help all of us further our individual quests to our Holy Grail of a Sound.

Enjoy and Explore

Fluid

Posted: 21 February 2005 at 7:54am | IP Logged

Many thanks for this M1 - it was good to re-read this, particularly the bit about compression settings