Recording - August 2016 - 54

as a processed track, then use the processed track for the final mix. The down side is
that, particularly with EQ, doing this while listening to the track by itself (using the Solo
button on the virtual console) means you won't be listening to the track in context.
I can remember times when I set up a track in Solo mode, then tweaked the EQ
and compression to sound absolutely perfect, only to have it fall flat when it became
part of the mix. Example: one time I mixed an acoustic rhythm guitar with no EQ (I
did use moderate compression), then decided it just wasn't working right in the mix.
When I added some upper-mid boost-a fair amount, maybe +5 dB-it sat in the mix
just right. When, just for grins, I soloed that guitar track with the EQ engaged, it
sounded perfectly awful. But in the context of the mix, it worked.
So the offline workaround, because it's done outside the context of the mix as a
whole, has its limits; it depends on your ability to hear in your mind's ear what the various tweaks will sound like in the final mix to come. That's hard. Fortunately, computers

What's going to happen to the
sounds of the other instruments in the
soundstage? To Dana's clarinet, not
much; it's panned mostly to the left
channel, and the compression is happening in the right channel. However
the two things panned to the center
(Will's vocal and banjo tracks) are a
different story. Sounds panned to the
center have equal level in both tracks,
but as the compressor turns down the
volume in the right channel (in
response to the fictional loud bass
note), the compression changes the
stereo mix's overall balance. As the

are improving quickly enough that pretty soon there won't often be a need to work
offline, because you'll be able to do everything at once without running into the
machine's limitations. In fact, depending on the complexity of the mix (and the computer being used), we're almost at that place now.
I mentioned compression; I was all set to apply a compressor to Shlomo's bass-not
to either of the bass tracks, but to the composite track that combined them in a group.
In a DAW, that meant inserting the compressor on the group bus rather than on an
individual track. As it happened, I didn't need to do that; Shlomo's touch is so even
and smooth that his signal didn't need compression. This makes him a rare and valuable bass player, and may help explain why his calendar tends to be crowded.
You can apply just about any effect to the individual tracks in a mix (including
grouped tracks); most DAWs offer some fancy stuff like flanging, tremolo, intentional
distortion, etc. Or you can apply effects to the mix as a whole, by inserting them onto
the main mixing bus rather than on individual channels. If you're using compression
on the whole shebang, however, here's something you need know about in advance.
This way and that
Let's say I put a couple of compressors on the two channels of the stereo mix for
some bus compression, making sure their controls-attack, decay, ratio, threshold-
are adjusted to the same settings. Since we're back in the realm of speculative fiction,
let's say further that Shlomo hits an un-Shlomolike note that's way louder than usual,
loud enough to trigger the bus compression. What's going to happen?
Take another look at Figure 1, the panning diagram. It's important to remember that
panpot stereo works by turning down the level of the channel you're panning away from.
Okay, so Shlomo hits his hot note, and since he's mostly panned rightwards, he
mostly triggers the compressor on the right channel. It does what you'd expect-it
compresses-which means, in effect, turning down its channel to cope with the tooloud rightward-panned bass note.
54

RECORDING AUGUST 2016

right channel gets softer, the vocal and
banjo get pulled to the left; when the
compressor releases at the end of the
bass note, and the volume in the right
channel returns to normal, the vocal
and banjo come back to the center. In
effect, anything panned to the center
will get pulled back and forth by the
action of the bus compressors.
It's fairly easy to avoid this problem
if you know what to do; in hardware
compressors, there's usually a way to
link two channels together, so that a
loud note in one channel pulls both
channels down equally. In hardware
compressors, this is usually done by
connecting the two channels'

Table of Contents for the Digital Edition of Recording - August 2016