Recording - March 2015 - 59

For every gain, there is a loss (turn, turn, turn)
The opposite of "gain" is "loss." In audio lingo, when the gain
is less than 1 (a fraction or negative number of dB), we call it
"attenuation." Mic inputs often include a switch to insert 10 or
20 dB of attenuation ahead of the preamp circuit to keep a
screaming vocalist or heavy-handed drummer from overdriving
the electronics.
Your mixer's faders are actually attenuators, though their dB
scale may go above zero because there's some gain ahead of
the fader. Faders ideally go all the way down to zero, but in
reality a little signal gets through. Sometimes a mixer specification will include attenuation at the bottom of the fader travel,
though it's often marked "-∞". [Infinity]
Oh, and one quick note about proper usage: a negative sign
on a gain means attenuation, so it's enough to say "20 dB of
attenuation". "Minus 20 dB of attenuation" is a double negative. Thank you.
Unity gain-a straight wire plus
To make a device useful in the real world, it may need a high
input impedance to avoid loading its source (a DI box, for example), or it may need a low output impedance to assure that it can
deliver its full level to the next device in the chain. If we want to
change only the impedance and not the gain, we can add a circuit called a buffer to the input or output, or both.

some faders are close to the top or bottom end of their travel, that's
a good indication that the input is getting close to clipping (fader too
low) or that you don't have enough preamp gain and you're amplifying the mic preamp's noise more than necessary (fader too high).
Most signal processors are line level devices that operate at
unity gain when they aren't doing any processing. Set an equalizer's gain controls for zero boost or cut and you'll get out just
about the same level as you put in, but when you turn the knobs,
you'll change the gain at certain frequencies.
Tubes and transistors and transformers-oh my!
Tubes and transistors are active devices-they usually amplify
unless they're in a unity gain circuit. A transformer is a passive
device, but it can have voltage gain, though at the expense of
current. If a mic preamp has an input transformer, the transformer likely has a 3:1 to 10:1 turns ratio, which provides 10
to 20 dB of the preamp's total gain. That's a bonus. It's the transformer's main job to provide the proper impedance match
between the microphone and the amplifier circuitry, and also
offer high common mode rejection. An output transformer is usually wound with a 1:1 ratio (unity gain) and is there to provide
a balanced output and ground isolation.
You might have read about an engineer patching a (usually
vintage) compressor into the signal chain and setting its controls
so that it's not compressing, just passing the signal through at

Gain vs. Frequency-a real-world case
Here's a lesson in why it's not sufficient to specify frequency response at a single frequency. In the late 1990s, when reviewing
a Mackie VLZ-Pro mixer, I measured 58 dB of mic preamp gain at 1 kHz, close enough to the 60 dB specification not to quibble.
However, I noticed that the gain was lower at around 100 Hz, and at 20 Hz, it measured only 53 dB.
Double check: No, the lowcut filter wasn't switched in, and since I was measuring the output level at the Insert jack, the equalizer wasn't in the measurement path. Backing off the gain by 10 dB resulted in a loss of only about 1 dB at 20 Hz, and with 40
dB of gain (2 o'clock on the Trim knob, about where most users, by nature, initially set it) the gain loss was less than 0.2 dB at
20 Hz-in my book, an acceptable deviation.
I was doing some documentation work for Mackie shortly thereafter, and since I had access to the engineering department, asked
if they were aware of the loss of gain at low frequencies. It turned out that they were, and that it was a conscious design compromise. By using a smaller-than-optimum-value capacitor at one place in the circuit, they were able to meet their cost budget and all
the parts would fit in the available space.
By the next generation (Onyx, VLZ3 and VLZ4 models), parts cost had dropped, they were able to fit a larger value capacitor in
the available space, and the mic preamp has full gain down to below 20 Hz. But the spec sheet still only states the gain at 1 kHz.
A buffer has a voltage gain of 1 (unity), so the same voltage
that went in comes out. However, an input buffer might also provide a differential ("balanced") input for a single-ended circuit.
An output buffer can provide a low output source impedance
(this is actually current gain), and is often used to provide a balanced output.
"Unity Gain" found common usage with the project studio mixers of the 1990s and continues today. Amplifier circuits have a
design center, an operating voltage range over which you get the
best compromise between noise and headroom. For best performance, it's necessary to raise the input signals to the design center level of the mix bus before they hit the faders, and keep them
close to that level when mixing, hence the term "unity gain."
There's usually a mark on the panel around 2/3 up from the bottom of fader travel that indicates unity gain through the fader.
With the channel fader, subgroup fader (if any) and the master
fader all set to their unity gain positions and EQ bypassed or set flat,
the signal level coming out of the fader will be the same as what
came out of the mic preamp. Of course, when mixing, the faders
will wander away from their design center positions-that's why we
have them. But if you find that when you have a balanced mix,

unity gain. What's the point? Even with no gain reduction, the
signal still passes through the input and/or output transformers
and/or a tube or solid-state buffer. That's where the "coloration"
that we like to hear from those devices comes from.
Nothing ventured, nothing gained
Understanding how gain affects the signal level from end to
end of a signal chain will help you choose the appropriate
equipment and adjust it correctly for the least noise and distortion. You need enough gain, but not too much. In the simplest
case, you need sufficient gain at the beginning of the chain to
bring your source up to your system's design center level, and
then keep everything beyond that somewhere around unity
gain, plus or minus a little for creativity.
In our next installment, we'll talk about a related characteristic,
sensitivity, and show how to determine from its specifications how
much gain you need for a particular source. See you then!
Mike Rivers (rivers@recordingmag.com) is a studio and live
sound engineer, gear design guru, and educator. Learn more
(lots more!) at mikeriversaudio.wordpress.com.
RECORDING March 2015

http://mikeriversaudio.wordpress.com

Recording - March 2015

Table of Contents for the Digital Edition of Recording - March 2015