Recording - September 2015 - 54
Real preamps can offer even wider gain ranges, ranging up to
+75 dB... not necessarily a practical number for music recordists, but
someone might use it in recording nature sounds, or doing surveillance. If the preamp is part of a large console or a console strip box,
this offers new ways to get into trouble.
If you set the preamp for too much gain, it will clip-that is, it will
go into overload, as it tries to put out more signal than it's built to produce. Figure 2 shows a classic clipped wave form (it's Will's banjo
track, actually, which I doctored to simulate preamp clipping).
What happens is that the positive and negative peaks of the audio
run up against the circuitry's limits. No matter what's on the input,
they can't put out more than their maximum, so they don't. The waveform looks like someone took a pair of scissors and clipped the peaks
off... which is why it's called "clipping". Clipping causes severe distortion, which makes the signal sound harsh and ugly.
fader-that setting is usually about 10 or 14 dB down from
maximum. If you set each fader to its mark, then turn the
gain on each preamp (often, on consoles, this knob is
called gain trim) to obtain the desired level at the output,
the console's design is such that the preamp shouldn't clip.
Setting faders in this way is a type of gain staging; we'll
encounter it again when we start mixing.
The second way manufacturers help you is by including
a clip light on the board-one for each input. You'll find
those on free-standing mic preamps too, on channel strips,
and on interface units. You should set the preamp's gain so
that the clip lights never flash. I've encountered "engineers", usually in live sound, who believed that the clip
lights should flash regularly on audio peaks. I'll say categorically: that's wrong. Preamps shouldn't clip, period,
because clipping sounds ugly.
All microphone preamps generate noise, which is added
to the microphones' inherent noise (I talked about that earlier). The manufacturer usually specifies this as equivalent input
noise (EIN), which is normally a large negative number. An
EIN of -128 dBu is excellent performance, -125 dBu is medium, and -120 dBu is decidedly subpar. If the manufacturer
Remember: if the preamp clips, the fader can only make
the clipped signal quieter-it can't unclip the signal!
"Hold on!" I hear someone saying. "The waveform isn't hitting the
maximum levels on the display, yet it's clipping. How can this be?"
To understand this, you need to check out Figure 3, an extremely simplified diagram of how signal flows through a console. (Drawings
like this are called block diagrams, and if you know how to read
them, they tell you a lot about what a console can and can't do.)
The signal moves from left to right: first through the preamp, with
gain set by you, then through the channel's fader, which reduces signal level by an amount also set by you, so that it can blend properly
with other signals. Then comes the rest of the console, which we'll
ignore for now.
If the preamp is set for too much gain, it'll clip; because the fader
comes after the preamp, it can't change what the preamp is doing.
If the preamp clips, the fader can only turn down a clipped signal-
it can't unclip the signal. What Figure 2 shows is actually a signal
that's been clipped, then turned down.
Since clipping sounds bad, it's a good idea to record so that nothing in the recording chain clips. Console manufacturers help you do
that in two ways.
Most consoles-from the fancy to the cheap 'n' cheerful-have
marks on the front panel that are the recommended settings for each
RECORDING September 2015
claims better than -130 dBu, just remember that the mic's
own noise (if it has a 150 ohm impedance) is -130.8 dBu,
so the manufacturer is probably fudging something.
A fudge that's actually legitimate is the "weighted" noise figure, in which the measured noise is filtered in a way that mimics the sensitivity of the human ear. In the USA this is commonly
called "A-weighting", and it actually makes sense, since it
attempts to quantify how much noise you'd actually hear.
Since weighted noise specs are inevitably quieter than
unweighted, they're almost universal in the world of commercial preamps, but that universality means that one
maker's specs will probably be comparable to another's.
So that's the basics of how preamps work and what their
specs look like. Next time, we'll delve into other useful features you'll find in most preamps, and consider how they
were used in our Ottoman Underground recording session.
See you then!
Paul J. Stamler (firstname.lastname@example.org) is a recording musician, engineer, educator, and collector of bizarre
vintage recordings, living in St. Louis.
Table of Contents for the Digital Edition of Recording - September 2015
Recording - September 2015 - Intro
Recording - September 2015 - Cover1
Recording - September 2015 - Cover2
Recording - September 2015 - 1
Recording - September 2015 - 2
Recording - September 2015 - 3
Recording - September 2015 - 4
Recording - September 2015 - 5
Recording - September 2015 - Contents
Recording - September 2015 - 7
Recording - September 2015 - 8
Recording - September 2015 - 9
Recording - September 2015 - 10
Recording - September 2015 - 11
Recording - September 2015 - 12
Recording - September 2015 - 13
Recording - September 2015 - 14
Recording - September 2015 - 15
Recording - September 2015 - 16
Recording - September 2015 - 17
Recording - September 2015 - 18
Recording - September 2015 - 19
Recording - September 2015 - 20
Recording - September 2015 - 21
Recording - September 2015 - 22
Recording - September 2015 - 23
Recording - September 2015 - 24
Recording - September 2015 - 25
Recording - September 2015 - 26
Recording - September 2015 - 27
Recording - September 2015 - 28
Recording - September 2015 - 29
Recording - September 2015 - 30
Recording - September 2015 - 31
Recording - September 2015 - 32
Recording - September 2015 - 33
Recording - September 2015 - 34
Recording - September 2015 - 35
Recording - September 2015 - 36
Recording - September 2015 - 37
Recording - September 2015 - 38
Recording - September 2015 - 39
Recording - September 2015 - 40
Recording - September 2015 - 41
Recording - September 2015 - 42
Recording - September 2015 - 43
Recording - September 2015 - 44
Recording - September 2015 - 45
Recording - September 2015 - 46
Recording - September 2015 - 47
Recording - September 2015 - 48
Recording - September 2015 - 49
Recording - September 2015 - 50
Recording - September 2015 - 51
Recording - September 2015 - 52
Recording - September 2015 - 53
Recording - September 2015 - 54
Recording - September 2015 - 55
Recording - September 2015 - 56
Recording - September 2015 - 57
Recording - September 2015 - 58
Recording - September 2015 - 59
Recording - September 2015 - 60
Recording - September 2015 - 61
Recording - September 2015 - 62
Recording - September 2015 - 63
Recording - September 2015 - 64
Recording - September 2015 - 65
Recording - September 2015 - 66
Recording - September 2015 - 67
Recording - September 2015 - 68
Recording - September 2015 - 69
Recording - September 2015 - 70
Recording - September 2015 - 71
Recording - September 2015 - 72
Recording - September 2015 - Cover3
Recording - September 2015 - Cover4