Recording - May 2018 - 64

Trust,But
verify

"monitor" setting to send the generator signal to the output
jack where you've connected the meter.
If your meter has a manual range switch, experiment with
it to understand how range relates to display resolution.
Increase the generator's frequency to see where your meter's
sensitivity drops off. Adjust the output level to see where the
"half digit" kicks in or drops off.
Once you've got the hang of using your test equipment, let's
use it determine the maximum output level of your interface.
This is the output voltage for a digital level of 0 dBFS. Knowing
this can help you manage your gain structure. First, to save your
ears and your monitors, turn off your speakers and unplug your
headphones. You'll be playing a track as loud as it can possibly
get, and that's mighty loud. Set the generator plug-in to 250
Hz and 0 dBFS out, or import a 250 Hz 0 dBFS WAV file to a
DAW track and start playing it. The frequency really isn't critical,
but any DMM worth using will be accurate up to 250 Hz.
Set the meter to read AC volts, and connect it to one of the outputs.
Pan the track with your test signal fully left or right to correspond to
the output you're measuring. This is important because most mixers
drop the track level going into the main mix by 3 dB or so when it's
panned to the center. By panning to one side, you'll be sending the
full-scale signal to the D/A converter. If everything is set properly,
you should measure somewhere between about 1.5 V for a typical
computer sound card and around 7 to 8 V for a "pro" interface.
Audio signal levels are usually expressed in dBu rather than
volts. There are online calculators to convert volts to dBu, or
you can calculate it using the formula
dBu = 20 x log (voltage ÷ 0.775)
But (there's always a "but")... if every interface was designed
properly, you could count on this measurement to be reasonably
accurate. However, I've found many interfaces and handheld

recorders with an analog output stage that can't quite handle
the output of the D/A converter at full digital level, and will clip
a signal that's a dB or two below 0 dBFS. Not only do you not
want to push the interface that hard, but unless you have a trueRMS meter, the meter reading will be inaccurate. We'll sort this
out in our next installment. (Hint: oscilloscope!)

THE WRAP
I really wanted to start this article by sending you to a hamfest or
on an eBay search for an old Hewlett-Packard 400 series AC voltmeter, and, while you're at it, an HP 200CD oscillator. (Figure 4)
Perhaps after working with less audio-specific test gear, you'll get
the bug. Even after 50 years, that rock-solid vintage test equipment
works just fine and can be had for as little as $25. But now that you
have some tools, experiment, and we'll look at input sensitivity and
gain measurements next time. Until then, happy exploring.
Mike Rivers (rivers@recordingmag.com) is an educator and
audio guru living and working near Washington, DC. The Editors
would like to thank Mike for the mental image of him testing
potatoes in his shop. To quote Dr. Magnus Pyke: "Science!"

The Sound Of A Potato

Last year, Paul Vnuk interviewed producer/engineer Sylvia Massy for Recording,
and she talked about using a potato as a "filter" to create interesting and
player-inspiring guitar tones. I was curious and a bit skeptical, so I took a guitar, an
amplifier, and a 5-pound sack of Idaho Russets to the shop, wired up a potato as
Sylvia described, and performed measurements to see what the potato was actually
doing in the circuit.
I began by making test leads consisting of #16 stranded wire with a 6d nail soldered
to one end. I could stab the potato with the nails and had convenient wires for
connecting it into a test circuit or to a guitar amplifier. Adjusting the spacing between the
nails would give me more or less potato, and I could easily swap potatoes. (Figure 5)

First, listen
On first listen, not only did it sound like a broken amplifier, but the volume was very low, even with the amplifier driven to its
maximum output. Rather than abuse my vintage Ampeg Reverbrocket, for bench tests I used a Mackie mixer with a DI input driving
64

RECORDING May 2018

Table of Contents for the Digital Edition of Recording - May 2018