Digital recording has been a boon to music in innumerable
and immeasurable ways. And the ubiquity of hi-tech
devices in our daily lives today—as well as the incredible rate at
which sophisticated features trickle down to the masses—has only
compounded the effect. Today, it’s affordable and remarkably convenient
for just about anyone to make high-quality recordings on
the run, in home project studios, and at rehearsals and gigs.
But while the possibilities are nearly endless with this new technology,
that doesn’t mean creating a great recording is simply a
matter of having a kick-butt computer, smartphone, or dedicated
recorder. Technology only gets you so far. Now, more than ever,
you have to know what you’re doing to get great results—because
it’s just as easy to make a bad digital recording as it is to make a
bad analog recording.
To that end, we’ve compiled this list of 10 must-know tips for digital
recordists to keep in mind no matter what platform or interface
you’re using. Armed with this knowledge, your guitars, your platform
of choice, and an impeccable set of ears, you’ll soon be cranking out
digital recordings that are clean, clear, fat, full, and ready to capture
the hearts and minds of listeners everywhere. Let’s dive in!
Turn it down. For the best fidelity when tracking, keep levels between −18 and −14 dB on the meters.
1. Turn it down!
In the analog tape days, the goal was generally
to get the hottest signal down to tape
as possible, just short of distortion. When
presented with high levels, analog tape
responds with soft compression and gentle
saturation that sounds great, especially on
drums and other percussive instruments.
But digital doesn’t work that way. It’s not
forgiving of peaks, and if you push it too
hard you run the risk of digital distortion,
which sounds awful. For this reason,
and to allow plenty of headroom for later
processing and mixing, it’s proven best to
record signals to digital at levels in the −18
to −14 dB range on the meters. This allows
plenty of headroom for peaks and for later
work on the tracks. Some audio interfaces
have built-in peak “over” protection, but
do yourself (and your tracks) a favor: keep
the signals in that −18 to −14 dB range,
hard as it may be to resist the temptation
to light up those meters! Your reward will
be cleaner tracks, better dynamics, way
reduced danger of digital distortion during
recording, and easier mixdowns.
sample rates matter. or not. Unless you have a compelling reason to use a
higher sample rate, 44.1 kHz is optimum for most productions.
2. Sample rates matter. 0r not.
The sample rate determines how many “snapshots” of your
audio signal are taken in a second. Sample rate determines
just one thing: the frequency response of the system—in other
words, the highest frequency the system can record. Today’s digital
gear is capable of working at 192 kHz sample rates (or even
higher), which provide frequency response up to 96 kHz—nearly
five times the highest frequency the human ear is generally
credited with being able to perceive. There may be situations
where those super-high frequencies are worth capturing. Some
“golden-ear” listeners claim to hear a difference when those
ultra-sonic frequencies are present, but this assumes the music is
being delivered to the listener on a DVD or some other medium
that supports sample rates over 44.1 kHz. Some recording
engineers also feel that certain types of input filters and plug-ins
work better at higher rates. But for most of us operating in lessthan-
perfect bedroom and basement studios, there are few reasons
to use high sample rates. A big drawback is that you have
to get that high rate converted down to 44.1 kHz to deliver it
on CD, which can potentially impact the audio quality. Another
is that high sample rates dramatically increase the size and computer
load for a project. If you have a full-on Pro Tools HD or
HDX system, that’s no big deal. But if you’re running a native
digital audio workstation (DAW), then that extra load can be a
problem. Personally, I record at 44.1 kHz if the project will end
up on CD or MP3. I rarely work at high sample rates.
(Top)Watch plug-in gain. Even if your mixer channel meters seem okay, it’s possible that you are overloading them
from inside one of your plug-ins. (Bottom) Choose where the color comes from. Many plug-ins that model or emulate the sound of vintage hardware
units—such as the Universal Audio Fairchild limiter plug-in shown here—can contribute nice colors to your mixes.
3. Bits matter.
The number of bits in a digital
signal affects its resolution and
dynamic range. Each additional
bit doubles the resolution of
the signal. The difference is
dramatic. A 4-bit signal has 16
steps of resolution to “measure”
and represent the signal—sort of
like using a tape measure only
marked in one-foot increments
to measure something. Moving
to eight bits gets us 256 steps.
Sixteen bits get us 65,536 steps.
And 24 bits jumps us to a whopping
16,777,216 steps! That’s
like using a ruler marked off in
100,000ths of an inch—we’re
talking super-fine measurements.
For this reason, it’s best to
track at 24 bit, then convert to
16-bit for CD or to MP3 as the
last step in the process. Having
all that resolution also allows us
to record at conservative levels
(see number 1, above) while still
having plenty of resolution and
dynamic range to make a great
recording. If you have a 24-bit
system and you record at −18 dB,
you’re still getting 21 bits worth
of dynamic range (each bit adds
6 dB of dynamic range) and
2,097,152 steps of resolution.
That’s plenty for excellent quality.
4. Watch plug-in
It’s so easy to drop plug-ins onto
a track to EQ, compress, or otherwise
process a signal. But be
careful—many of those addictive
little software processors
can change the gain or level of
the signal, which often leads to
clipping (distortion) that’s easy
to miss in the heat of mixdown.
Be sure to double-check the
meters and clip indicators inside
each plug-in to ensure that no
overloads are slipping past you.
Boosting the gain in the plug-ins
too much can also force you to
have to pull down the channel
and master faders later, which
isn’t good gain staging and can
impair the audio quality.
5. Choose where the color comes from.
Analog tape and processors color your recording signal in a slightly different way—that’s part of their
beauty. Today, we can emulate a lot of those colors with modeling software that recreates the sound of the
original gear, including the coloration it provides. By selectively choosing the modeled processing you use,
you can carefully add analog-style coloration exactly where you want it and where it can be most useful.
Drums, for example, often benefit from analog-tape-style or analog-compressor-style coloration. Delicate
acoustic guitar, on the other hand, may not fare so well if you hit it with a highly colored processor.
6. Get it right at the source.
Just because we can do so many things with digital
audio, doesn’t mean we should—it doesn’t relieve
us of the responsibility to capture great sounds at
the source. Life is so much easier, projects come
together so much faster, the process is far less painful,
and the final product is so much better when
the original tracks are in tune, are tonally solid, are
recorded well, and aren’t plagued by background
noise. “Fix it in the mix” is just as much of a negative
with digital as it was with analog.