FOR MUSICIANS ONLY
Hurricane Sessions
Mixdown Phase One
December 6, 2008 -- I've
already described my initial approach
to dealing with the individual tracks, the point of which is
to make each one have enough gain and compression to give each
instrument or input the strongest possible clean signal without
overtly changing the way anything sounds. The only difference
is that now there 16 to 24 tracks to deal with for every song.
Step 1, as always, is to make
a backup. I always feel more willing to take chances with things
if I know that I can totally destroy a track and go back a couple
of steps without effecting anything else.
Despite the simplicity of using
a plug-in, I prefer to do most of the compression manually, even
if I'm dealing with 24 tracks, which at least one of the seven
songs managed to require. This is just the beginning of my bizarre
personal approach to digital mixing. So this is my next step
and it takes forever just to get through the drums. Moreover,
I do a big chunk of the manual compression for all the tracks
before I even listen to any of the songs. I may add some automated
compression as a secondary step, but usually there are still
a few peaks holding the rest of the track down, so I'm going
to go over it again.
There are seven songs. I spent
two full days on this alone. This is a pretty hardcore quirk,
but it has made all the difference in the way my recordings sound,
particularly over the last three years. They're much fuller.
I think it adds a little of what vinyl enthusiasts call warmth.
These two days are tedious and boring, but you can see that much
obvious correction just by looking at the tracks. I'd rather
adjust the level of a dozen or two individual sound waves instead
of running things through a processor and squeezing the life
out of it.
After this, it's time to do
another backup. I still haven't done much of consequence to anything,
but I've taken a long time to do it. I don't want to have to
repeat this process. So this will become the new reference backup,
which we'll maintain even after later changes have happened.
Before I do, however, I'm going to go inspect all the tracks,
make sure they're all labeled properly, trim the excess from
the beginning and ends, do a "Consolidate Selection"
for each track to remove any dependencies on exterraneous tracks
and make sure the names of the files match the names of the tracks.
This lets us thin out the file list before we really get started,
saving some possible later confusion.
I'd say that the most significant
lesson in my conversion to digital was to do as little as possible
to the sound of the tracks until you're at least this far along.
No EQ to speak of. It's like the total opposite of doing live
sound, where you have to make everything happen at once and there's
no going back for a do-over. With recording, you have time to
be meticulous. You can look at those peaks in relation to the
rest of the track and make judgement calls over one or two dB
for individual sound waves, in order to boost the overall level.
In a live setting, what I spent
two days going through so far would happen in about three minutes.
I'd have the band run through one or two songs, set a starting
point for levels, tweak a few EQs and let them take a break until
showtime. So far, we've only set the initial gain.
Drums
Even though the same drums
were used, in the same setting, without altering anything between
songs, we don't want them to sound exactly the same in every
song. So as I go through and listen to just the drum tracks,
I'm going to listen to the first one closely, add plug-ins for
EQ. Then I'm going to zip through all the songs and to the same
basic set-up for each one. This doesn't address the "sounding
exactly the same" issue, but using the plug-ins means they're
not etched in stone.
The kick drum mic I used (Nady
DM80) is relatively flat from 150 Hz to 1k Hz. It needs a notch
filter to boost the 340-360 Hz area, which is where the tight
thump of a rock kick drum lives. While the "perfect"
frequency is dependent on the size of the kick drum and its corresponding
resonant frequency, it can be softened a little or accentuated
by varying the exact setting. I'm also going to add in a couple
of dB at 6.2k and above, plus a little at 3k for clarity.
The snare drum has an ugly
low-midrange ring. Before I simply hunt it down and kill it,
I'll move the emphasis by boosting the high end frequencies.
This will be further helped when we strip the low end out of
the hi-hat mic, and accent its highs as well, so the two tracks
are somewhat dependent on each other.
The toms all benefit from a
2-3 dB cut at 440 Hz to reduce overtones and a 2-3 dB boost at
4.8k. At least one song has a floor tom track that is more or
less empty. Manny never hit it. There is also the question of
whether to strip out all the background from the tom tracks,
just leaving the few instances where he hit the drum. Or maybe
just do it with automation later. For now, I'll leave it.
The overhead mics I'm going
to leave alone for the moment, for much the same reason. I'm
not listening to enough of the song to understand their value
yet.
Submixes
My next step is to build submixes.
I add stereo auxiliary inputs for this. How many will depend
on the song, but once determined, a bus is selected as the input
for each submix.
Notice that the vocal submix
and the vocal effects share the same source. I'll apply the effects
100% wet via a plug-in, but keeping them on a separate track
means that use of the effect will not lessen the presence of
the vocal.
I'll use a submix for the vocal,
even if there's only one. Simply feeding a mono track into a
stereo submix fattens it up and this alone is the key to making
sure you don't bury the vocal track. Unless you want to.
Part two of this (shown below)
is to go to every track and switch its output from the main mix
and route it to the appropriate submix. In the example, I'm leaving
the bass in the main mix, as there is only one bass track. We're
not going to submix it with the guitars, so it can stand alone.
The only drawback to this is
that if you try to solo a track that is in a submix, you have
to solo the submix, too, which will have you scratching your
head the first few times you run into it.
|
 |
December 7, 2008 -- So
far, I made a set of mp3s to capture the raw version of the songs
and I have an additional backup of all the sessions after the
first, most time-consuming clean-up work. In the last step, we
did preliminary drum EQ and built the submixes.
After I've listened to the
drums for a while, across the various songs, and not without
a few tweaks, I'm going to keep the EQ settings. So I check the
plug-in, write down the parameters, remove the plug-in, go to
the edit page and apply the EQ in a destructive edit. Now it's
there, all the plug-ins are gone and it is time to start getting
picky.
I'm going to make a second
generation backup of the sessions as well, just to give me another
stage to go back to.
At this point, it already seems
like I'm being kind of anal about making back-ups. In the end,
I'm only going to save a couple of the generations. But while
the process is ongoing, yeah, I make a lot of damn backups. As
I said earlier, it's because I like to do destructive edits.
I rarely actually go back to early generations to pluck back
out a raw track, but the few occasions that it has happened make
it worth it. I was a moron when I originally did the Hayden's
Wall album. I didn't do submixes and, as a result, some of
the vocals suffered a little. Having those original tracks let
me go back and give it a better treatment, which is the version
I sent to retail.
I already offered up the raw
version of "Cry." So, for comparison, here's how the
sound has progressed so far.
|
