Owners Manual

VU METERS
Two toggles are used for the VU meters. One is used to select
whether the VUs display Input Level (after the Input Pot), Output
Level, or the OPTO Limiter Gain Reduction. If in BYPASS the
OPTO GR the meters drop out rather than sit on zero. A VU meter
showing OPTO GR seems to be a bit of a standard and the time
constants and ballistics are a good match, even if the VU does not
show every drop of gain reduction. The LED meter also can display
Opto GR or the total gain reduction.
The second toggle is a pad or attenuator for the VUs in the Input or
Output modes and has no effect on GR mode. The 3 position switch
gives 0 (no attenuation and calibrated to +4dBu), -3 dB and –6 dB.
While it is very unusual to get an attenuator for VU meters on
standard rack processors, it is a feature we have been building onto
mastering consoles for many years. A regular VU meter would be
continually pinned in the red without the attenuator (which may
have a disturbing effect on some clients). There are several reasons
for this, some pre-dating the trends of heavy squashing. Compare
individual acoustic tracks at DFS, to a mix at DFS, and the mix will
generally look hotter on the VUs. Mixes tend to have more average
energy than individual tracks. Both mixing engineers and mastering
engineers often (usually) compress a mix, which also increases the
ratio of average to peak somewhat. For example in recording,
typical peak to average ratios are 16 to 20 dB, but in mastering 14
dB or less.
The SLAM! is a Swiss Army Knife and is intended to be used for
much more than just mixes and mastering. Because it can severely
reduce the ratio of average to peak levels, and because most of us
now reference levels to digital full scale (a peak reading), it follows
that without the attenuator, the poor VUs would be ‘in the red’ much
of their life. We caution that the –6 position can mislead one into a
situation where the audio is really too hot, or too squashed. On
mastering consoles, the most often used VU pad is –4dB (a hint).
METERING GENERALLY
The VU meters and the SLAM! peak meters will never agree and
maybe the SLAM! peak meters are a little 'off' from the peak meters
on an external ADC or DAC. What is going on? Which ones do I
trust? How can I calibrate them to look the same?
These questions have been around for as long as there have been
meters. VUs should match well with other VUs because there is a
comprehensive list of standards and qualifications to meet to be a
true VU meter. Part of the specification are 'dynamic characteristics'
which describes "The pointer shall reach 99 on the percent scale on
0.3 second, overshoot less than 1.5% (.15dB)". This essentially
makes the VU meter act with "approximate RMS" response and our
ears have an "approximate RMS response". Ahhhhh loudness.
Peak Meters are much faster and are supposed to catch events less
than 0.0001 Sec (compared to the VU's .3 Sec) which means that
transients have a much bigger influence on peak meters. One
problem is that our eyes won't be able to adequately see a 0.1 mSec
flash or even 1 or 10 mSec so the designer has to stretch the duration
of the leds or pointer. In a sense, this exagerates transient contribution,
but at least looks good. There can also be a peak hold dot that adds
a digital stretch to the duration and allows us the luxury of blinking
or looking away once in a while. We have that in mode 2.
In pro audio there are few measurement standards (or even de-facto
standards) that exist and peak meter calibrations and digital converter
I/O levels are more prime examples. With digital peak meters that
are married to digital converters, one at least expects that an 'over'
in the A to D over would light the top red LED. Wrong - the better
meters use 4 samples over and the best allow that number to be user
set. 4 samples over is about the threshold of where we hear a clip.
Makes sense to me.
The SLAM! LED meter is not digital, it is an analog meter with a
micro-processor. It is not reading the AES-EBU data streams, and
just measures 6 analog inputs plus 1 signal from the optional A to D
(an over that's equivalent to 4 samples at 48K). It can be adjusted for
different analog sensitivities with an internal trim pot but this is not
guaranteed to 'match' other peak meters nor was it intended to. The
SLAM displays approximately 1 dB per segment, except near the
bottom which are much bigger steps to show signal present. Other
meters may show 2 dB or 1/2 dB steps or be dB scaled on a curve.
Our solution to allow some sort of matching with other peak meters
(like the ones on your favorite A to D), is to give you a way to set
where the color changes occur. The transition from green to amber
is settable, and for mode 2, also the transition from amber to red. This
can be done from the front panel without ever removing the unit
from the rack or unscrewing the top (mode 3).
In order to provide some useful indication of A to D clip or overs,
any green LEDS turn red, which is hard to miss. What is interesting,
is that because the meter is looking at analog levels, one can see how
many dBs the peak went beyond A to D clipping and this is also a
good indication of how audible the clip was.
So to answer the question "Which one should I trust?" the best
answer is "all of them with some interpretation and a grain of salt".
The VU is best to show apparent volume or apparent GR, the
SLAM! peak meters show analog peaks, headroom in the SLAM!,
and momentary GR that may or may not be audible, plus shows
optional internal A to D clipping. External digital peak meters on the
recording device should accurately display available headroom and
clipping there. None of them is 100% accurate, nor implicitly
perfect, and because each may have different response characteristics,
might look different especially after a peak and the rates that dots fall
and/or hold. If this is what you would like to calibrate, sorry.
Regarding calibrating exactly how many dBs it takes to hit digital
full scale on the A to D, we didn't (and couldn't) put a little trim pot
somewhere handy. Why not? The whole front panel is a bunch of
level controls or controls that affect level and there ain't no detents.
In other words this isn't just a basic A to D that does nothing else.
You don't have trim pots, you get a slew of big knobs and 4 limiters
to get optimum levels, but not finely calibrated. The other reason is
that the A to D front end is a transformer, not a bunch of op-amps.
For the DAC there are trims that can be used to calibrate the back
panel 1/4" balanced +4 dBm outputs. We set them so that DFS = 8.0
volts RMS (-16 dB from DFS = 1.27 volts or +4.3 dBm), which is
one common standard but there are others (-14 to -20). The DAC
signal is reduced 6 dB from the -16 DFS standard if you select the
DAC as the SLAM! input. With so many tracks and mixes squashed
these days, it allows a better sweet spot for the SLAM!. You can
always turn up the INPUT 6dB if you need. With the INPUT at 12:00
a -10 DFS should give 0VU and the 4th LED should light. DFS will
pin the VU and the LED will hit about half way up (LED #14).
14