The Light of Day

So you want to make an album? (part 12)
KeithHandy posted in Producing, So You Want..., Tools on April 16th, 2007

Installment 12: Standard sound recipe 101.

I recently uploaded some “unplugged”-type acoustic performances to YouTube. I did some originals (I recommend watching at least one of them before reading the whole post) and some covers; as of this writing I haven’t edited and uploaded the covers yet. While this is a fairly popular thing to do nowadays, I decided that I wanted better than average sound. So, I set up studio-quality conditions, using a few simple guidelines that I can spell out for you here, which should kind of establish some helpful basics for everything you ever record.

1. Quiet the room.

Duh. But seriously, if you have heaters, fans, air conditioners, or dehumidifiers running all the time, you probably cease to notice them. My heater is weird, and if you supposedly turn it “off”, it likes to throw temper tantrums at random intervals; so I befriended the circuit breaker. Since I was shooting video too, I “quieted” the room visually by using a simple black sheet for a backdrop, which had the side effect of helping to define my mental focus and establish that I was doing a “performance”.

2. Quiet your mind.

Of course, before your pre-performance mental psych-up, you should have the technical stuff all up and ready to go. Once it is, and you’ve done some mic and level testing and so on, put that out of your mind and give yourself a few moments to switch to performance mode. Take a short walk, have a small snack, visualize your audience, and/or contemplate your reason for wanting to do that song.

3. Quiet your voiceless bilabial plosives.

A windscreen or “pop filter” is very important if you plan to record any vocals. The violent low-pitched “thud” you hear every time the letter “P” is uttered is very difficult to remove once you’ve already recorded it. It may not be so noticeable on small speakers, but for anyone with a subwoofer, it will sound like someone is trying to kick their door in. (Maybe that’s a good thing.) No, the little foam thing that fits on the end of the microphone is not good enough.

If you’re not vain, you can make your own filter with nylons stretched over any kind of hoop. Set the screen at least a couple of inches out from the mic, and keep your mouth at least a couple of inches away from the screen. You should be able to feel it suppressing short gusts of air when you try to blow through it, but it will be otherwise acoustically transparent. Make it two-tiered for extra-anal pop prevention.

4. Choose your transducer.

I’ll confess — for what I do, I should have condenser mics, and I haven’t gotten around to buying one. Last I checked, Behringer made some that started at only a hundred bucks each. Alas, the mics I’m using are Shure SM-57s, a popular example of dynamic mics, which are more ideal for the stage than the studio. They don’t have quite as crisp and “open” of a high end as a condenser would, but they are more directional, so at least they have the advantage of picking up less extraneous sound, like the scuffling of pet mice, or the mating calls of drunk twenty-somethings just outside your window. You can record decently with dynamic mics, but again, you could record even more decently with a condenser.

By the way, the above-linked wikipedia article is impressive; I’d never heard of liquid microphones or laser microphones before!

5. Give yourself head.

Er, that’s supposed to say headroom. Damned union typesetters and their rigid break schedules. “Headroom” is the amount of leeway available for you to get even louder without distorting — the distance above your loudest recording level before you hit the ceiling. This is especially important if you record digitally, since overloading a digital signal does not make for a pleasing, “musical” kind of distortion. Make sure that at the very first stage, wherever in the chain your signal converts from analog to digital, that when you are loud, you have at least six to ten more decibels above that, just in case.

If you have three different places in a row where the level of your input can be adjusted — say there’s a knob where you have your mic or instrument plugged in, and then some kind of overall recording volume for the computer, and then a record level on the recording software, you want to set all but the first adjustment in that chain to either “100%” or “0 dB”, which both mean the same thing. Then, adjust whatever is at the very beginning of that chain to get the ideal level going to the hard drive — again, -6 to -10 dB when you get loud. Maybe you will only be hanging out around -20 dB most of the time, and it might seem like it should be louder, but in digital this is definitely okay. You want to be free to play and sing without worrying about distortion.

6. Be naked and unashamed — initially.

Sometimes effects are part of a performer’s sound, like when a guitarist uses distortion, wah-wah, and a zillion other fun toys. But the “finishing touches” — reverb, compression, and equalization — should not be committed to the track as you’re putting it down, because you will want to have control over these when you mix. Most recording software allows you to hear your sound through special effects while recording, even though what you’re actually recording is just the raw sound. In other words it keeps the recording stage separate from the effect stage. This way, after you record it, you can increase, decrease, or remove the effect; whereas if the effect was part of the recording itself, you would not be able to change it; you would only be able to pile more effects onto it.

7. Perform.

If you have the above items taken care of, you should now be able to hit “record” and dive right in. I recorded continuously for about an hour and a half, so I was able to forget about the computer and just focus on playing, singing, and enjoying myself. I considered wearing headphones, but it would have been distracting. If you don’t create a situation where you can liberate yourself mentally from the technology, you won’t be completely in the performance. You will need to do some mixing later, regardless; but mixing an inspired performance is a pleasure, while mixing an uninspired performance is a nightmare.

8. Normalize.

Now that the element of chaos has subsided, and you’ve recorded everything a bit on the quiet side, it’s safe to bring it up to a more useful initial volume. Most software has a “normalize” effect that you can apply to a track. It’s nothing fancy, just a volume boost that automatically scans the whole clip first to see how much it can safely increase it by. Go ahead and normalize each track to 100% and/or 0 dB. (On the software I use, the waveforms shrink and grow interactively as I adjust the level of the clip, so I can just do this visually.) This has the added benefit of making the waveforms easier to see.

Note: if you are using a stereo input to record two different things simultaneously, say guitar and voice, the left and right will have to be split into two separate tracks so you can work with them independently. How to do this depends on your software.

9. Compress individual tracks… a little.

Almost anything with that “pro” sound has at least a bit of compression on it. Not to be confused with data compression which is used to create smaller files, dynamic compression is the nearly-instantaneous smoothing out of amplitude over time. A friend of mine who was less familiar with this term told me he was surprised that it made the sound so “big”, when the term “compression” in his mind made him think it would “sound smaller”. The reason for this is that it gives sound more overall energy relative to its peaks, so you can create the sense of a louder sound while the maximum remains the same. I won’t go into the role this plays in the loudness wars, because that’s a whole article in itself. But suffice to say, a conservative amount of compression on the individual tracks will give them a little more energy, and make them easier to mix together.

10. Add reverb to whatever needs it most… as little as it needs.

If you’ve ever heard anyone talk about using reverb to cover something up, they probably use too much. Sure, there may be times when there’s a good aesthetic reason to put something in a cavern or a cathedral. Even in those cases, you shouldn’t go that far on every instrument; there should be something more “dry” in the foreground to contrast it against. Remember that reverb is sound. When you use it, you are adding a sound to the mix. This might seem like a dumb thing to point out, but the point is, if a song is getting cluttered, reverb can compound the problem. In most cases a vocal should have just enough reverb that you can hear a difference when you switch it on and off, but not enough to be more than barely noticeable otherwise.

11. Equalization… huh! What is it good for? Absolutely nothing! Except…

Use EQ when you need EQ. Again, this sounds obvious and dumb, but the point is, you usually don’t. There are a lot of different kinds of filtering effects that manipulate the highs, lows, and mids in a wide range of precisions and proportions. These include graphic equalizers, parametric equalizers, FFT filters, “scientific” filters, highpass, lowpass, and bandpass filters. They’re all about slicing the audible spectrum into bands to be adjusted separately. A “graphic equalizer” would be the most intuitive for the beginner, as you’ve probably already played with one on your home stereo.

If you have an instrument or sound that doesn’t use the full spectrum, it’s sometimes good to filter unused frequencies to minimize noise. For example, you can easily reduce treble on a bass guitar to remove hiss without taking anything important away from the sound of the instrument. Likewise on higher instruments you can reduce bass to remove rumble. (I rolled off a little of the “boominess” of the acoustic guitar, and left my vocal un-equalized.) On a very mid-rangey instrument like a distorted electric guitar, you can usually roll off a little at both ends.

On acoustic instruments, you generally want to be sparing with any kind of equalization, as you can easily make them sound worse — especially when your ears are getting tired. Try to get the sound right in the first place with good microphone placement. If you must use EQ creatively to solve a problem with the sound, try to cut bands rather than boost them, or you may run into problems with unpredictable spikes in volume at the boosted frequencies.

12: Chip away at everything that isn’t part of the statue.

How hogwild you want to go with this is a matter of personal preference, but if your voice doesn’t even start until five bars into the song, you might as well erase that track up until the point where it starts. Your guitar intro will thank you for letting it be all the cleaner. Just be careful whenever you cut a clip; if you’re not careful there will be a “click” at the point of snippage. There are two ways to avoid this: either always do a fade-in or fade-out at the edit point, which can be as short as a few milliseconds — or always cut the clip exactly on a zero crossing, which is exactly what you would expect it to be: a point where the waveform is crossing over the “zero” line.

13: Take care of the mix while the master is away.

Another thing cool new software lets you do these days is to place “mastering effects” onto the whole mix so that you can get a slick, cool, “produced” sound right away. This generally amounts to “multi-band compression”, the bastard lovechild of compression and equalization. It’s like using equalizers to split the sound into bass, midrange, and treble, and then running each of those three bands into its own separate compressor. It sounds f’n great. HOWEVER, you should try to get your mix sounding good without it first.

The fact that you’re going to apply some compression to the overall mix is one of many good reasons to be sparing with it on the individual tracks. Why would I recommend using it on individual tracks and on the overall mix? What’s the difference? A little compression of any kind on a full mix — and multi-band is all the better for this — will just help to blend and cement the various instruments together a little, and make them sound more like parts of a whole. At this end of the project, you’re shaping the whole puzzle, whereas previously you were shaping the pieces.

If you are a noble soul, you will use these tools for primarily for aesthetic reasons. You want that little “boost of life” that a mastering effect or plugin can give you; as far as the loudness wars go, you can still be a conscientious objector (and damn well you should be). But, that said, when your song comes up, you don’t want someone dropping the shopping bag with the eggs in it just to fumble with the volume on their iPod — so that bit of volume increase doesn’t hurt.

So, okay, Jimi, you say you want your recording to be wild and artistic and unconventional, so why am I teaching you how to bake white bread? Because!!! Because you can’t color outside the lines if there are no lines to color outside of. Because you can’t push the boundaries if there are no boundaries to push. Because you can’t fuck the man if there’s no man to fuck. Just trust me — someday you too may be old, boring, and grumpy, and just want a vocal to sit in a goddamned mix. You’ll thank me for this someday.