As I continue my latest whimish adventures in recording, for some reason I’m awash with lessons taught me by all my engineering friends. Thought I’d share this one today: the technical background for one of engineering’s best loved axioms, “Garbage in, garbage out.”
Garbage in, garbage out is a pretty simple rule of thumb. Also known as “You can’t polish a turd,” and a whole bunch of other sayings including one old one about silk purses and sow’s ears. Basically, if what you start with is crap, it will likely end up as crap, and if you don’t do some things to maintain quality from the beginning in your creative process, you are very likely to kick yourself later.
To keep yourself out of garbageland and eliminate the do over while you record, apply some simple gain staging techniques.
Gain staging is basically conditioning the signal you are recording so that it is strong, clean, and powerful from the beginning of the process. In this way you “stage” the gain – by either boosting it or cleaning it up so that you do not have to work with something wimpy or crappy later in your recording process. You absolutely do not want your project littered with tracks that are noisy, lack power, or simply don’t deliver enough solid gain to hold up to other tracks. If you get all the way to the mix with a bunch of badly staged tracks, you are – absolutely, without a doubt – going to want to shoot yourself. Every newbie recording engineer has experienced this at one time or another and it totally blows.
Signal – what we call the incoming electrical impulses that your recording system receives – is after all, the physical building block that your recordings rely on. So without good staging, signal becomes less of a foundation in your recordings and more of a liability.
The reason is something engineers commonly refer to as “signal to noise ratio,” sometimes abbreviated SNR or S/N. The signal to noise ratio is basically the ratio of signal power to the noise power corrupting the signal. It can also be thought of as the difference between the noise floor and the reference level. The reference level being variable, but for this blog, let’s just say it’s you screaming into a microphone.
If there is more noise than signal, obviously, you have a problem. And it’s our goal in any studio to make sure that the noise part is mostly kept out of signal. I say mostly because there is going to be some amount of noise present in anything you record. You may have noticed that if you hit record on your digital audio workstation while recording nothing, that track may still have some noise in it. That’s because systems have a base level of system noise, particularly if you are working on a computer.
If you’re a non-math person like me and prefer to spend your time listening rather than learning complicated formulas, what you really need to know about signal to noise is what to listen for on your system.
A good place to start is by simply plugging a mic or instrument cable directly into the input of your I/O box. If this is the way you normally record, then just go ahead and do what you usually do and enable a track to record on. Now go ahead and turn the trim (or gain) knob on your I/O box or mixer and sweep from zero to max. You should immediately hear a range of noise introduced. You’ve also found the single biggest noisemaker in your system.
Mixers and I/O boxes have trim/gain knobs because they have an onboard preamp. These “pres” enable you to plug instruments directly into the mixer or I/O, if necessary. The catch is that the onboard pres – even on expensive I/O boxes or mixers – usually suck. Because think about it – it’s just one piece in the channel strip. Inserting incredibly awesome pres on your I/O box is too costly to make the box profitable, and so therefore, right out of the box, you get something that is usable, but not great.
The problem with this in the home recording studio is that new engineers may get the idea that they really don’t need a separate instrument or mic pre and that the best way to get a stronger signal is simply to turn up the gain knob. But if you’ve been using your gain knobs as your primary input control for all your signals, you’ve been unwittingly introducing a ton of noise into your tracks. This would be an example of bad gain staging – where the signal is not being conditioned or strengthened at all before hitting your system, which means you end up having to artificially boost it using the not so great pre at the front of your system. Not a good plan of attack in gain staging.
Alternately – a good preamp that you purchase separately and has both input AND output gain control is what you want in order to limit noise in your tracks. People who are just getting into recording ask me all the time if a preamp is really necessary. The short answer: absolutely. Without one, you can achieve some mediocre results. With one, you can take your home recordings to a whole different level. Even having a preamp in the cheaper price range is better than not having one at all. Because they allow you to get good clean input from your mic or instrument while separately adjusting signal output to your system. When it comes to gain staging, this double level of control is really what you’re going for. And having the external pre allows you to eith
er turn your gain knobs way down or bypass them altogether, further limiting noise.
If you’re working with microphones, you are probably really going to need to experiment to determine proper gain depending on what you are recording. If you’re recording vocals, the process may be as simple as setting your preamp levels and hitting the button, however, if you’re mic-ing up your amp, there are going to be a whole lot of factors that affect your staging, including microphone type and directional settings, proximity to your amp, and noise coming off the amp.
Since we’re going for low noise, start with your amp. Listen to it both by itself and through your system. I listen through headphones because phones tend to boost the frequencies in the mid and upper ranges where noise most often sits, so it makes it very obvious to identify. Eliminate whatever noise you can with your amp whether it’s because of bad shielding, or an old cable or a cranky connection to your pedalboard.
Next test your microphone at different ranges in front of the amp and at different positions. Positions will be different for different mics, so you’ll have to figure out based on your setup where/if noise is being introduced this way. You can check out mic technique diagrams, but keep in mind that all environments are different and experimenting and listening is a great way to find the sweet spot working with your particular tools.
Sometimes it helps to baffle the amp and mic using either professional baffles, or your own homemade versions. I’ve used blankets, pieces of wood, closets, adjoining rooms and sound panels that usually hang on my studio walls. This technique – depending on the mic – allows the mic to pick up just the amp and no reflective sound or outside noise. This is particularly useful if the amp is in the same room as your computer and all your other equipment.
Finally, back at your system, make sure the signal is clean and clear when it comes into the channel, and isn’t peaking. Even in digital, you want your signal to be “hot” without overload, which results in clipping, popping, or an absence of sound. On my audio input meter, I will typically seat the signal just below peak at its most dynamic moment. Stay away from the red, but don’t be afraid to get close. If your meters consistently register below the 0 point, you are going to have weak tracks that you have to push the faders to max on later.
From the MIDI side, gain staging still applies. If you’ve played a track, but notice random or weak velocities in your MIDI note information, don’t be afraid to boost them. You can either manually plug new ones in at the problem spots, or apply an increase to all velocities. Either way, make sure that your MIDI instruments are putting out good solid signal and you don’t have to pull the fader up to max just to hear them in your mix.
And here’s my all time best gain staging trick: I personally hate using compression while recording, particularly on my own vocals. But there are those moments when the dynamics of a track are all over the map and you have to do something to keep the track from going from whisper nothing to blowing your head off.
If you are recording someone else, take the time to learn their performance a little so that while you are recording you can manually ride the output of the preamp. This allows you to achieve consistently solid gain staging at your system while the dynamics are pumping without having to use a compressor. It takes some practice to do this on your own vocals, and is obviously not possible if you are playing yourself, but if you record often with others, it is a trick worth trying together. It works great for vocalists who are incredibly dynamic.
So, those are the basics of good gain staging. Experimenting often and totally getting it down in the early stages of recording will save you a lot of hair follicles later. Don’t ever expect to be able to fix a weak, poorly gain staged track “in the mix.” It won’t happen.
You will simply end up caught in the time consuming and frustrating reality of garbage in, garbage out.
UPDATE: An engineer friend in LA sent me this awesome link this morning. For the more advanced engineer readers, but a technique that is well worth considering, particularly if you’re going to be mastering outside your system. Have at it!
Originally posted 2010-09-27 19:19:18.