So a friend was making a short film for an art college project, and asked me to contribute a voice-over-narration-type-thing.

I thought I’d make a blog post on how I went about it.

I have no proper recording equipment; the only microphone I have is a cheap-ass ten-euro-at-Argos mic whose intended use is for shouting at people on webcams. I’m plugging it straight into my computer’s microphone jack.

This does not bode well for sound quality. Those mics are of the minimum acceptable standard for anything, and the mic inputs on computers are notoriously shite - you really need a proper mic preamp if you want a good recording. I suspect this latter point is the source of the noise problems to come.

First comes the question of the script itself. Said friend wrote it, and it was good, but I gave it a few tweaks, because fairly often what seems ok on paper just doesn’t work in speech.

This is an odd phenomenon. On the one hand, having to read something aloud seems to highlight any minor awkwardness in flow or sentence structure - long sentences with big sub-clauses usually just don’t work, for example, because by the time you’ve got to the end of the sentence, the listener can’t remember how it began - but on the other hand, the spoken word is often completely incomprehensible in writing.

I heard somewhere that this first penetrated the public consciousness during the Watergate scandal, interestingly enough. Transcriptions of tape-recorded conversations appeared in newspapers and the like, and people were bloody confused by them - they were just jibberish. How could elected politicians be so inarticulate?

But of course, we’re all like that in speech, because our social interactions are largely understood through body language and inflection in the voice, rather than what we say. I find it very frustrating when people don’t recognise this; my mother, for example, thinks you’re not listening unless you’re staring straight at her.

Anyway, I tweaked a few things, and started trying to get a good sound from the microphone.

Mic placement is everything. The position of the mic, X and Y, is important, because depending on where it is it’ll pick up a different tonal spectrum from your voice. This is something the human ear seems remarkably deaf to. I guess something in the audio processing part of our brain compensates. But microphones can be a lot more senstive to this than you’d think. By placing the mic below you, looking up at the roof of your mouth, you get more high frequencies reflected into it. By placing it higher, you get relatively more bass; this is sort of counter intuitive, as you’d expect the proximity to your chest to increase bass - it doesn’t seem to. Closer to your nose, maybe a more nasal quality.

Also, the distance of the mic from the source is important. For some reason, mics tend to exaggerate distance; if you’re six inches away, you sound like you’re three feet away. If you’re three feet away, you sound like you’re speaking from the wrong end of a cave. To get a good, hard, radio-presenter sort of quality, you need to be really close.

This creates another problem. When you speak right into a microphone, breath noise becomes a real issue. Again, we don’t hear these things in speech, but the microphone sure does. If you’re not careful, every plosive T and P, and every sibilant S and fricative F smacks the mic with twice as much volume as the vowels, and makes for a completely bizarre sound, likely distorting your recording in the process.

Oh, that’s the other problem; you need to control the input volume somehow.

Signal-to-noise ratio is important. The noise, any hiss or hum in the background, is constant, so if you can increase the volume of the main signal - your voice - you can make the noise seem relatively quieter. So the louder the input signal, the better. Until it distorts. The problem here is that you’re sitting on an unfortunate knife-edge; too quiet, and the noise seems too prominent; too loud, and you distort the recording with a big crunch. There are various clever studio tricks to cope with this. I don’t have the ability to use any of them.

I ended up sitting my guitar amp on the table, the mic on top of that (oh I’ve no useful mic stand by the way…), and keeping it a few inches from my mouth. I leaned in and out a little as the voice got softer and louder, respectively.

Next came the actual recording itself.