Technical Background - Behringer EURODESK SX3282 User Manual

20
EURODESK SX3282 User Manual
10.6 Recording Levels
When recording to digital, it's a good idea to keep the recorder's peak meters
below 0 dB. Most (not all, esp. samplers) read 0 dB with some headroom left.
This is because, unlike with analog, the onset of digital distortion is as sudden
as it is horrible. If you really want to take your recording level to the limit
(and fully exploit 16-bit digital's 96 dB dynamic range), you'll have to do some
calibrating. How to do it? Well, you could run a tone at 0 dB from the mixer and
use that as your DAT reference. But your DAT may be way under its maximum
input limit. Probably a better way to work out just how hard you can drive your
recorder is to incrementally increase the record level until the onset of digital
distortion, subtract, say, 5 or 10 dB, and never exceed that level. Engage "peak
hold" on your recorder before recording if you want to confirm that you haven't.
When recording to analog, the tape machine's VU meters should show around
+3 dB on bass, but only around -10 dB for hi hat. Although analog distortion
is more like compression at modest overload levels (often desirable on bottom
end), higher frequencies cause saturation even at modest levels (an unpleasant
"crunchiness"). Also, VU meters tend to progressively under-read above 1 kHz,
due to their sluggish response time. Hi-hats should read about -10 dB on a
VU meter, as against 0 dB for a typical snare drum, and +3 dB or more for a
kick drum.
Peak meters read more-or-less independent of frequency. Aim for 0 dB recording
level for all signals.
10.7 Track Sheet
When laying out channels for recording or mixing, try to be sensible.
Keep tom-toms together, etc. Work out a scheme that suits you & stick to it.
A common order is: kick drum, snare, hi-hat, tom-toms (as the audience sees
the kit), cymbals (ditto), bass, guitars, keyboards, other instruments, vocals.
From session to session and gig to gig you will soon know where you are without
ever having to look at a track sheet.

11. Technical Background

11.1 Mixing
11.1.1 Equalization
In the beginning, EQ was an instrument for removing unwanted frequencies,
or compensating for imperfect microphone response curves, or bumps in a
studio's acoustic. It was a corrective device. Tamla Motown turned that notion
upside down in the sixties with the novel idea that you try to find for each
instrument a characteristic frequency not shared by the other instruments in the
mix. Then you whack up its gain. This makes individual voices punch through a
mix in a slightly unnatural but exciting way.
In general, corrective EQ usually involves broadband (slope) contouring,
together with narrowband notching of unwanted resonances. The narrower the
notch or "Q", the less the total signal will be affected.
Finding bad resonances is made easier by first frequency sweeping in
BOOST mode.
"Motown" EQ is achieved by applying boost in a fairly broadband way.
The broader the band, the more musical but less instrument-specific the effect.
Applying boost over a narrow bandwidth will sound "honky". For sounds which
require drastic corrective EQ, it is advisable to have a couple of channels of fully
comprehensive parametric equalization in your rack. (You can always bounce
tracks through the outboard EQ, freeing up the unit for the next task).
Check out the ULTRACURVE PRO DEQ2496, a superlative digital stereo equalizer
and much, much more.
For "advanced equalization", EQ might be applied to a signal as follows:
First, trim the LF and HF shelves to achieve the required slope or "loudness".
Now use a parametric EQ band to boost the most significant frequency for each
instrument or tape track. Over all channels, if two or more of these frequencies
coincide, then you might have to settle for second best in some cases, if you want
to achieve optimum separation in the mix. Really nasty frequencies will need
notching out.
A vocal signal can be enhanced by applying a significant boost in the 12 kHz
region or higher, above the nasty sibilance region. This is especially effective if
you've got a de-esser patched post-EQ.
◊
Use the low-cut filter to tighten up channels in a mix: maybe
remove it only for the bass, kick drum, toms, tablas, didgeridoo and
other deliberate subsonics (when recording classical music ignore
this advice).
With the LF set to boost and the low-cut switch activated, you have got a peak
response rather than shelving at the bottom. Good for tight but deep bass.
Remember EQ contouring can be done with cut as well as with boost.
Cutting away the top and bottom, then pushing up the gain is equivalent to
mid range boost! EQ is not a one way street!
Always reset a channel's input gain (or external devices' output level)
after altering the amount of desk EQ cut or boost applied.
11.1.2 Gain Optimization
PFL (Pre-Fader-Listen) is the way to set a desk level. Master aux send levels are
fixed at unity gain. As the mix progresses, more and more channels are likely to
be sending to effects via the aux buses, and it's best to PFL all sends just before
setting up for the final mix.
Outboard reverbs etc. should all be made to work hard. There's no point in
having an 85 dB dynamic range if the input meter of your reverb is barely
flickering. On the other hand, digital distortion is not one of the nicer noises
around. You'll have to rely on your ears to detect digital distortion, since different
outboard processors calibrate their meters differently.
If you hear distortion, turn down the input on the FX unit, and turn up the desk's
aux return input.
99 times out of 100 distortion in the aux send > FX > aux return loop will come
from the FX unit (FX gain too high), and the same goes for a high noise level
(FX gain too low).
Analog multitrack tape should be driven quite hard, since its dynamic range
(without noise reduction) is likely to be 20 to 30 dB worse than other elements
in the recording chain. Try to record bright. You can always mix back duller.
Brightening up an off-tape signal will bring up the level of tape noise.
With digital tape or hard disk you have plenty of dynamic range, and treble
pre-emphasis is not often necessary. Just don't let the signal distort!
When mixing or recording, keep the channel fader levels around or below 0 dB.
If you do find the faders creeping up or down, apply a suitable offset over all
channel faders, and try to control your bad habit in future!