Manley SLAM! - Manual (Page 14)

‘Hyper-Compression’
The word ‘Hyper-compression’ is a word mastering engineers use and
was coined by Lynn Fuston (Mastering Web Board, DSD vs 96/24).
For many people it implies the idea of limiting and multi-band limit-
ing and normalizing and squeezing every last drop of apparent volume
possible onto a stereo mix but most mastering engineers would pre-
fer not to. It has become almost a contest and everybody wants to be
louder than everybody else. Record companies expect it, or ask for it
and sometimes demand it. Mastering engineers are expected to do it
and are given mixes so brutally squashed that they can’t get any louder
anywhere without just clipping and distorting.
‘Maximizing’ level also implies ‘minimizing’ dynamics and transients.
Dynamics and transients are one of the few available ‘elements’ in
sound and music, with the others being pitch, duration and waveform.
Very primitive music was just log drums and stretched animal skins,
or mostly just transients and dynamics. It’s not hard to make ‘music’
with just transient information, but can you do it with just pitch (like a
technician’s oscillator), or the one of the other two elements? Dynam-
ics and transients are musical elements and not the enemy.
Maximizing levels very often results in a CD that can exhaust the lis-
tener before the first song is over. It can produce an aggressive in-your-
face constant barrage that might not be appropriate for every project,
every song, and every artist. It might be like most effects, best used as
an effect and where it is appropriate, sometimes full tilt and sometimes
lightly. Super-squashing might also be going out of fashion, and we see
a trend where producers are avoiding it.
Maximizing is best done at the mastering stage rather than during mix-
ing, usually. We’ve heard these stories many times, “The A&R guy
(or producer or artist) demanded that the engineer use a certain box
to maximize the mix, so he did, but he also supplied the mastering
engineer with an alternative version without that box. The mastering
engineer did his best with both versions, and everybody preferred the
results from the ‘raw’ tape and hated the maximized mix, and in the
end the one from the ‘raw’ version was loudest anyways”. Never heard
the opposite story. Why? Mastering engineers generally have the best
tools, the most tools, and the right tools for the job, and a ‘pre-mas-
tered’ job often robs the mastering engineer of the opportunity to apply
those tools, and experience, and abilities. If you do use the SLAM! on
a mix (before it gets professionally mastered) you should have a ver-
sion without it and/or a version with light limiting. They also like 24bit
masters, and a little headroom (like -3 to -6 dB below digital clipping)
and lots of accurate labels/notes. Thank us later.
Maximizing generally does not help the song sound any louder on the
radio or TV. Perhaps you know that they are old hands at that game and
have compressors, 10 band limiters, 4 band limiters, full range limiters
and clippers strapped across the audio chain all the time and have had
for 15 years. Everything comes out the same volume - as loud as they
can make it. Things can get silly because their boxes were set to work
on ‘normal’ mixes and sometimes they get goofy when given super-
squashed songs and songs that have lots of stereo or out-of-phase info.
For them, width is bad, mono is good, it transmits further, gets bigger
audience share, bigger ad bucks.Car CD players like over-compressed
material and many now build a bad compressor into the car stereo, so
this is covered too.
So, the SLAM! is another “GO-LOUDER” box, but with a warning
label. We are building guns, not pulling triggers. As always, use your
ears, judgement, and taste. Be careful with this cannon.
Hints
Everybody knows that you should make those A to D peak meters go
as hot as possible, digital full scale, but NEVER clip. Well, we have
two urban myths in that statement. Often enough, the next thing
after the A/D (filters, plug-ins, processors & EQs) requires a few dB
of headroom and they might have a nasty distortion complete with
aliasing if given digital full scale. It is caused by a little ripple in
the pass band of the filter that actually can add a little bumps across
the spectrum. The ‘better’ the filter the lower those bumps will be,
but they add up. Most filters need between 1 to 3 dB of headroom.
Some MPEG encoders need 6 dB. Don’t feel compelled to hit DFS
if you are working 24 bit and going to master later. Resolution will
be fine and a little headroom may be a nice thing 6 dB below DFS.
Save almost DFS levels for mastering 16 bit/44.1 CDs that need
every bit.
Yes, most early A/Ds sounded horrible when pushed into clipping
and clacked, barked and complained loudly. Most modern A/Ds can
be pushed a bit ‘over’ into clipping without obvious distortion and a
few (especially older Ultra-Analog based ones) have a pleasant clip.
You can certainly get a hotter mix, and more ballz with a little clip-
ping (keyword being little) but our advice is to choose where and
how to clip carefully. Usually analog clips better than digital and
usually tube units clip better than solid state, but lots of tube boxes
don’t sound good at all clipping (including some of ours). So, it
requires experimentation, (on your time), careful level scaling, and
careful listening to highs, lows, peaks and noise. If you do it right,
maybe, make a master with level to rival the top mastering engi-
neers, or do it a bit wrong and ruin the project, and have something
you regret for a long time. It is not easy or automatic, and boxes that
claim to do it usually don’t.
A large part of the trick to getting a hot loud mix is to watch the
VUs. The idea is not to see how far the needle can bend to the right,
but how still and immobile you can get it and still have it sound like
music. Make that needle just sit there while mixing before you limit
and then you just need to use a few dB of limiting and loud it will be.
This is not so easy to do. Limiting individual tracks and sub-groups
makes it easier. Start with the hottest tracks, which are usually vo-
cals, bass and drums. This applies for basic limiters, multi-band lim-
iters and de-essers. Each of these are harder to use effectively on a
mix than an individual track or sub-group. On a mix, anything hot
can trigger them and they affect everything. If you de-ess a mix, try
not to mess up high hats, acoustic guitars & careful EQ settings.
De-essing a lead vocal is relatively easy. Limiting a mix will prob-
ably seem to affect the drums first, because in a typical mix drums
are the source of most of the transient spikes. The initial attack or
transient gets pulled down (spikes you probably liked during track-
ing) but also everything else gets pulled down for that instant. So
drums begin to sound more distant and feel pulled back in the mix
(and you spent an hour fine tuning that part of the blend), they begin
to sound dull (because the transient contains the bulk of their highs),
and reverbs and room mics seem to get louder. When limiting hits
vocals, some good things and some questionable things can happen.
Plosives (Ps, Bs Ds Ks) might get hit and alter carefully sung pronu-
ciations, but sometimes, after EQ, plosives need some taming. Some
singers hit high notes a lot harder, and fader-riding, compressors
or limiters are needed (hopefully, in that order). In the end, if indi-
vidual tracks and sub-groups are limited first, then, not only is final
limiting and mastering easier, but also it is easier to mix, as well as
to create dynamics in the mix. By the way, vintage recordings rarely
used a buss compressor - its a fairly new trend. If you do limit the 2
buss, watch out for quiet sections and the drum balance in the mix,
and use those ears.
14