The lightning flash with arrowhead
symbol, within an equilateral triangle, is
intended to alert the user to the presence
of uninsulated "dangerous voltage"
within the product's enclosure that may be
of sufficient magnitude to constitute a risk
of electric shock to persons.
TO REDUCE THE RISK OF ELECTRIC SHOCK, DO NOT R EMOVE COVER. NO USER SERVICEABLE PARTS INSIDE.
REFER SERVICING TO QUALIFIED SERVICE PERSONNEL.
TO REDUCE THE RISK OF FIRE OR ELECTRICAL SHOCK,
DO NOT EXPOSE THIS APPlIANCE TO RAIN OR MOISTURE.
TO REDUCE THE RISK OF FIRE, REPLACE ONLY WITH
SAME TYPE FUSE. REFER REPLACEMENT TO QUALIFIED
SERVICE PERSONNEL.
CAUTION
RISK OF ELECTRIC SHOCK
DO NOT OPEN
WARNING:
THIS APPARATUS MUST BE EARTHED
1. INTRODUCTION
The Ashly DPX-100 combines a 15 band graphic
equalizer and full function peak compressor/limiter in a
single rack space product. Both equalizer and compressor/limiter can be used as stand-alone processors, or can
be automatically chained together with a back panel
switch.
The graphic equalizer section uses professional
quality 25mm faders for maximum resolution and long
life. Ashly graphic equalizers use "Q" enhanced Wein
bridge filters with interleaved summing for constant "Q",
low ripple, and minimum filter interaction.
The Ashly compressor limiter circuit was designed in response to the need for unive rsal peak-sensitive automatic gain control (AGC) devices with
exceptional audio performance and rugged durability.
The result is a wide-bandwidth, ultra-low-distortion, low
noise VCA (voltage controlled amplifier) which is versatile and highly listenable.
Premium components are used throughout the
DPX-100, and computerized automatic assembly equipment verifies that each component's electrical specifications are within tight tolerances before becoming part of
the circuit assembly. Each finished unit is then tested
twice before leaving the factory, guaranteeing you a
worry-free, professional product for many years.
Please read this instruction manual thoroughly
before operation so that you may realize all the features
and benefits the Ashly DPX-100 has to offer.
The exclamation point within an eqilateral
triangle is intended to alert the user to the
presence of important operating and
maintenance instructions in the literature
accompanying the device.
2. UNPACKING
As a part of our system of quality control, every
Ashly product is carefully inspected before leaving the
factory to ensure flawless appearance. After unpacking,
please inspect for any physical damage. Save the shipping carton and all packing materials , as they were carefully designed to reduce to minimum the possibility of
transportation damage should the unit again require packing and shipping. In the event that damage has occurred,
immediately notify your dealer so that a written claim to
cover the damages can be initiated.
The right to any claim against a public carrier
can be forfeited if the carrier is not notified promptly and
if the shipping carton and packing materials are not available for inspection by the carrier. Save all pac king materials until the claim has been settled.
3. AC POWER REQUIREMENTS
A standard IEC-320 AC inlet is provided on the
rear panel to accept the detachable power cord shipped
with the unit. Units distributed within the United States
are preselected for 120VAC, 60Hz and should be plugged
into a standard NEMA 5-15 3-wire grounded AC receptacle. Most units distributed outside the US are
preselected and labeled for 240VAC, 50-60Hz and are
shipped with the appropriate power cord.
The DPX-100 will perform normally from 95 to
125 volts AC. An internal line fuse is used. In the event
of fuse failure, refer to a qualified service technician for
servicing. Power consumption is less than 20 watts.
Model DPX-100
Gra phic Equ alizer
Compressor/Limiter
10K
+15
+10
+5
0
-5
-10
-15
Gain
dB
+6
+4
+2
0
-3
-9
-
∞
Range
HPF
20Hz
OutInOut
±6dB
±15dB
EQ
Sig Clip
4. MECHANICAL INSTALLATION
The DPX-100 mounts in a standard 19 inch
equipment rack. The mounting screw threads vary with
different rack manufactures and you should refer to your
rack instructions for proper hardware. An oval head or
flat head screw with a plastic countersink washer is preferred to protect the finish of the DPX-100 under the screw.
This unit is housed in a rugged steel case and
will tolerate moderate abuse. However, for road systems
which may be dropped or otherwise subjected to extreme
forces, we recommend some rear support for the chassis
to prevent bending the front panel the front panel when
these forces occur.
For installations where it is desirable to protect
the front panel controls from tampering or accidental
misadjustment, use the Ashly security cove r. Installation
is simple and does not require removal of the equipment
from your rack. See your Ashly dealer for details.
5. EQ CONTROLS
5.1 EQ Boost/Cut
The individual equalization faders adjust the
boost or cut at each filter frequency. By adjusting a combination of faders, an overall frequency response can be
developed and the physical position of the f aders will gi v e
an approximate visual indication of this response.
5.2 Gain
The gain control adjusts the overall gain of the
equalizer when the EQ switch is “in”. It is generally used
to compensate for level changes due to the equalization
process, but can also allow the equalizer to adjust overall
system level and gain structure. Overall gain with this
control is from +6dB to -∞.
5.3 Range
This switch selects the operating range of the
individual equalization faders to either ±15dB or ±6dB.
The ±15dB settings should be used when much equalization is needed. The ±6dB setting allows finer resolution
on the fader settings.
5.4 HPF (Hi-Pass Filter)
The high-pass filter can be used to supplement
the frequency response achieved by the bandpass filters.
It’s function is to “roll off” low end response to eliminate
Ashly Securi ty Cover Installation
subsonic interference like wind noise, floor rumble, and
boomy microphone pops. The high-pass frequency is fixed
at 20Hz with an 18dB/octave slope.
5.5 EQ In/Out
This switches out the boost/cut controls and EQ
gain adjustments. This way, the effect of any equalization can be compared to a “flat” response. This switch
does not defeat the high-pass filter.
5.6 Sig and Clip LEDs
The signal present LED comes on when a signal
greater than -20dBu arrives at the EQ input. The clip
LED indicates any EQ signal level exceeding +19dBu.
nal level to the VCA circuit. It is always active, so
switching out the limiter function has no effect on this
control. Used in conjunction with the input/output level
meter display, this control is useful for setting up optimal
system levels. This control should normally be left at
"0" to achieve accurate threshold calibration.
6.2 THRESHOLD
The threshold control has a range of -40dB to
+22 dB, allowing applications from low level compression to high level limiting. The threshold control de-
termines the audio level above which gain reduction
occurs. When the threshold LED comes on, that means
that gain reduction is beginning to occur, due to input
signal peaks exceeding the selected threshold in dB.
6.3 RATIO
This control determines the resultant change in
output level to changes in input level for all signals above
threshold. The numbers printed around the ratio con-
trol are calibrated in db and indicate the increase in
input (above threshold) required to produce a 1db increase in output. This can be expressed conveniently as
a ratio. If the output remains constant no matter how high
the input level, we have an infinite (∞) input/output ratio. It should be remembered that the ratio control has no
effect on signals which are below threshold.
There is a common but incorrect notion that limiting always implies the use of an infinite ratio. Although
there are times when an infinite ratio is desirable, there
will be situations where infinite, or “hard”, limiting action is neither appropriate nor necessary. In fact, it should
be noted that an infinite ratio setting is likely to cause
noticeable side effects in the sound, and may not be usable on programs where subtle control is desired.
Clip
-6
-10
-20
0
-3
-
+3
+6
+10
+15
+20
∞
dB
Output
Inpu t
.5
Sec
1
2
3
5
.2
10
15
.1
Power
6.4 ATTACK TIME
The response of the compressor/limiter to signal
levels abov e threshold is further defined by the attack time
control. Attack time is the amount of time it takes to
attenuate the output level after threshold has been
reached. For v ery fast transients, such as hand claps, snare
drums, or other percussive sounds, a fast attack time is
usually desirable so that the limiter can respond in time
to control the peak level. On other types of program material, a slower attack time may be preferred. An abrupt
attack may, on some material, “square off” the top of a
waveform, producing a distorted sound . The DPX-100 provides continuously variable attack times from 200 microseconds to 20 milliseconds.
6.5 RELEASE TIME
Another parameter which affects compressor/lim-
iter performance is release time, or the time required to
restore system gain to normal after the input signal
has fallen below threshold level. Again, proper release
time will depend on the type of program material being
processed and the way in which the limiter is being used.
When subtle limiting is desired, slow release
times are often chosen to avoid condition referred to as
“pumping” or “breathing”. This occurs when overall gain
is modulated up and down by repeated peaks which are
followed by quieter intervals. If the release time is set
too fast, then the overall level will jump up and down,
producing an objectionable and unsettling effect. Note
that, in some cases, an individual track or channel which
seems to be pumping may sound acceptable when heard
in context of a complete mix.
A unique feature of Ashly Compressor/Limiters
is the incorporation of a double release-time constant.
When a conventional compressor/limiter is adjusted for
slow release times, transients such as mic “pops” may
cause a severe reduction in gain followed by a slow fadeup, making the action of the limiter very obvious. With
the double time constant, release from gain reduction after a brief transient is always fast, with a slower release
after a sustained overdrive.
6.6 OUTPUT LEVEL
Output level control is provided to fully cut or
restore up to 18 dB of system gain. For unity gain, set
the control to 0. NOTE: When the compressor/limiter is
switched out, the output control still functions.
6.7 IN/OUT SWITCH
This switch enables you to quickly hear the compressor/limiter in or out of the audio chain. When the
switch is in the out position, all limiting and compression controls and functions are bypassed, with the exception of the gain and output controls, which continue to
function as straightforward level controls.
INPUTS are Active B alanced.
OUTPUTS May Be Wired
B a la n c e d O r Un ba la n c e d .
3
2
XLR
1
Fem ale
Shown
(-)(+)
Output
6.9 INPUT/OUTPUT METER SELECT
While the Gain Reduction display accurately represents the action of the limiter, comparing input to output levels in real time is somewha t more intuitive, and is
made simple using the input/output meter select switch.
The input meter takes its signal just after the gain control, and will indicate input signal level regardless of output levels or limiter settings. The output meter display
takes its signal from the actual output of the unit, so every control that affects the output will also have an effect
on output meters. Used in conjunction with the gain reduction meters, input/output meters prove to be an extremely useful diagnostic tool when working with system
dynamics and level control.
7. CONNECTIONS AND CABLES
Com
TIP = Detec
(Use
RIN
6.8 THRESHOLD/GAIN REDUCTION DISPLAY
As soon as the threshold level is reached, the yellow LED illuminates. Depending on how far the input
level rises above threshold, successive red LED’s will illuminate, indicating gain reduction. Gain reduction can
best be described as the difference between input level
and the resulting change to output level. For signals
below threshold, there will of course be no gain reduction, that is, a 10dB increase in input will yield a 10dB
increase in output. For signals above threshold however,
output level will increase only to the extent that the ratio
control allows. With a high ratio, say 20 or so, it will take
20dB of increased input level to increase output level by
1dB. With a gentler ratio of 3:1, input signals above
threshold will be “gain-reduced” at the output by 1/3. In
other words, with threshold set at 0dB, a signal peak at
+12 dBV that is 3:1 compressed (ratio at 3) will produce
only +4 dB (12÷3) at the output, and 8 dB of gain reduction has occurred (12 dBV input minus 4 dBV output=8
dB reduction.)
7.1 Balanced vs. Unbalanced Audio Connections
Balanced signal connections are preferred in pro
audio applications because of their improved immunity
to induced hum and noise. A properly shielded and wired
balanced input stage on any audio product will reject most
unwanted noise (RFI, EMI) picked up by the cable, as
well as minimize ground loop problems. Therefore it is
always advantageous to use balanced connections when
running signal more than ten or fifteen feet, although particularly noisy environments may require that ev en short
patch cables be balanced.
Unbalanced connections are used mostly for short
distance, high level signals (0dBu nominal). Most external EMI noise pick-up will be masked under the noise
floor of the signal, assuming there is little or no gain following the unbalanced signal. If a gain stage does follow
a signal, or if externally sourced noise persists, use balanced connectors.
The DPX-100 uses two different audio connector types. 1/4" TRS (tip-ring-sleeve) phone jacks, and
three pin XLR connectors will allow interfacing to most
professional audio products. Ashly TRS balanced connections use the tip as (+) and the ring as (-) signal, with
sleeve used for ground. Ashly XLR connectors use pin 2
Out
Output
7.3 Chain
The chain button on the back panel allows the
output of the equalizer to be fed directly to the input of
the compressor/limiter, with no external cable required.
When the chain switch is in, the input connectors to the
compressor/limiter are removed from the circuit, while
the graphic equalizer outputs remain functional.
(+) and pin 3 (-) with pin 1 ground. Inputs are 20KΩ
active balanced using precision 1% metal film resistors ,
7.4 Detector Loop - Ducking
outputs are 200Ω "pseudo-balanced", which means they
have balanced impedance with a single-ended signal
source, and can be wired balanced or unbalanced. When
possible, we recommend balanced connections between
all components in your system.
sert DETECTOR PATCH point which can be used as a
"ducking" input, or in conjunction with an equalizer to
produce frequency-sensitive limiting. Various uses of the
The DPX-100 compressor/limiter has a TRS In-
detector patch are discussed under TYPICAL APPLICA-
If inputs are used unbalanced, the signal should
TIONS.
be on the (+) connection and the (-) connection must be
tied to ground, or signal loss will result. While a mono
phone plug used as an unbalanced connection will automatically ground the (-) ring of the jack, XLR's will not
8. TYPICAL APPLICATIONS
automatically do this, so attention must be given to proper
wiring.
By itself, the 15 band graphic EQ is useful for
general tone control, feedback control, and room reso-
Tip (+)
Ring (-)
Sleeve (Gnd)
Stereo Phone Plug
used for balanced
nance correction. The compressor/limiter provides many
solutions where dynamic signal level processing is required. The combination of graphic EQ and compressor/
limiter allows for additional applications, such as a full
range speaker processor, 70 volt distributed system pro cessor, mixing console channel insert, and frequency sensitive limiting, to name a few. In most cases, the DPX-100
Tip (+)
Sleeve (Gnd)
Mono Phone Plug
used for unbalanced
should be the last device before the power amp or crossover, or right before a recording device or transmitter.
8.1 GRAPHIC EQUALIZER APPLICATIONS
PUSH
Input
XLR pins are
numbered
on the
connector
insert.
2 = (+)
3 = (-)
1 = (gnd)
Audio C onnector Types
XLR Male
XLR Female
General Tone Control
The graphic equalizer is a very useful device for
general tone shaping because it is intuitive and easy to
adjust. The visual reference provided by the slider position gives an approximate idea of the frequency response
generated, with the lower frequencies on the left and
higher frequencies on the right. To use the power of an
equalizer effectively, you need to translate your idea of
the tone you want to produce into a range of numerical
frequencies. This is simple after a little practise. Here are
a few references which are useful for starting points:
- Very low bass (the “wind” in a kick drum,
almost felt as much as heard -40Hz-80Hz.
Try using these starting points as a guide when
you want more or less of these types of sounds. Adjust
by ear from there. It is always a good idea to remember
that a little equalization usually works out much better
than a lot, and that there are many audio problems which
can not be solved with equalization alone.
Feedback Control
A graphic equalizer can be used to provide some
control over moderate feedback problems, but does not
have enough flexibility or resolution to handle sev ere situations. You will achieve the best results when you can
eliminate one or two feedback points by setting one or
two sliders for no more than a 6dB cut. Often you can
find a feedback point by boosting sliders in succession to
determine which frequency ranges contain the feedback
modes, and then cutting those ranges. Be very careful in
this process to avoid explosi ve feedback and possible system and hearing damage! If you find feedback points
with many equalizer bands, remember that cutting every
band may not help (all you will do is reduce system gain).
The combination of a graphic equalizer for tone control
and a parametric equalizer (such as the Ashly PQX-571
or PQX-572) for feedback control is highly recommended.
Console Channel Equalization
Many mixing consoles provide only simple equalization for individual channels. If your console has channel inserts, you can patch your graphic equalizer into a
channel that’s being used for something important and
use it to tailor the sound of this channel exactly the way
you want.
Large Room Equalization
Large rooms tend to suffer from multiple reflections with long time delays, long reverberation times, and
“ring-modes”, all of which lead to reduced intelligibility
and a generally “muddy” sound. As sound travels long
distances through the air, high frequencies are attenuated
more than low frequencies. In general, large rooms benefit from some low frequency roll-off, high frequency
boost, and attenuation of ring mode frequencies. As in
the case of feedback control, a graphic equalizer can help
reduce an isolated ring-mode or two, but a tunable narrow-band equalizer such as a parametric is more effective here.
8.2 COMPRESSOR/LIMITER APPLICATIONS
As the functional name implies, a compressor/
limiter can be divided into two basic categories, limiting
and compressing. When used as a protective device to
prevent audio levels from overloading systems such as
tape recorders, power amplifiers, speakers, or transmitters, it is generally referred to as a limiter.
It may also be used to create special effects and
unusual sounds for recording and musical performance
by deliberately reducing the dynamic range of a signal,
creating a much louder or fuller sounding signal without
increasing the loudness peaks, in which case it is referred
to as a compressor.
The Limiter As A Protective Device
The DPX-100 compressor/limiter section provides fast and accurate gain control for the prevention of
sound system overload due to unexpected transients.
Sound system distortion is usually the result of amplifiers running out of power, in which case nice round waveforms turn into harsh sounding squared-off waveforms.
Looking at it from the perspective of the speaker diaphragm, this means that, whereas in normal operation the
diaphragm is required to accelerate, slow down, smoothly
change direction, and accelerate again, distorted operation requires an instant acceleration, instant stop, a change
of direction, and instant acceleration again.
Since speaker diaphragms are subject to the laws
of physics, they won’t take this kind of punishment for
long. The diaphragm may shatter, or its voice coil may
overheat. In addition to the damaged caused by sustained
overload, the speaker may also be damaged by occasional,
one-shot high level overload, for example, the sound of a
microphone falling face-first onto a hardw ood floor . Even
if this type of transient doesn’t destroy a speaker outright,
it may damage the speaker surround in such a way as to
cause mechanical abrasion and future failure.
Alternatives For Sound Installations
To install a compressor/limiter in a sound system using a passive crossover, insert it between your mixing console output and the power amplifier input. For
systems using electronic crossovers, there are two ways
to use a compressor/limiter. It may be inserted between
the mixer output and the crossover input, in which case it
will act on the entire audio frequency spectrum. Alternately, if the limiter is inserted between an output of the
electric crossover and the input of a power amp, it will
only affect a specific band of frequencies.
Recording
The Ashly limiter can be used to prevent tape
saturation in analog recording. Also, with modern trends
toward inexpensive digital recording, it remains necessary to protect against input overload. With digital recording, the information stored on tape, hard disk, optical
disk, etc., is either a 1 or 0, so actual signal level on the
tape is not the concern it is with analog recordings, in
fact it is not even a user controllable parameter. What is
of concern however, is the signal level applied to the A-D
(analog to digital) converters. If clipping occurs at the
converter input stage, the resulting distortion is most unpleasant, and will be recorded digitally as if they were
part of the original audio signal, forever mixed with the
audio. To prevent converter distortion while preserving
the extended dynamic range of digital recording, look up
the max input level of your recorder/conve rter and set up
the limiter as follows:
1. Set Gain to 0.
2. Set Threshold to 2-3 dB below max conve rter
input.
3. Set Ratio to 10.
4. Set Attack to 2 mS.
5. Set Release to .2 Sec.
6. Set Output level to 0.
If you are exceeding threshold frequently, your
input signal is probably too high and should be turned
down. Of course, eve ry situation is different, so experi mentation before final recording is always a good idea,
but this is a good starting point.
To obtain a gentler limiting action a t the expense
of some dynamic range, decrease the threshold to -15 and
the ratio to 3-5. This is also a good starting point for
analog recording.
Broadcasting
Compression has long been used as a
tool to make an audio signal appear louder. A
good example is in broadcasting, where com-
peting stations with identical transmitters and power attempt to sound louder than each other. Since they are all
restricted with respect to maximum audio level (modulation), their best tactic is to squeeze the dynamic range of
their programs to just a few dB. The audio output level
of the station virtually never changes, and the listener
perceives this continuous high-lev el sound as being louder
than the same material in an uncompressed form. Although
both compressed and uncompressed programs reach the
same peak levels, the compressed signal stays near peak
level more of the time, and thus sounds louder. This technique makes the broadcast more intelligible over ambient noise, and increases the geographical area over which
the broadcast is audible to the listener. Additionally, this
compression technique is extremely useful for FM and
infrared transmission systems for the hearing impaired.
8.3 Special Effects
Compression For Feedback Contro l
A common ritual in sound system set-up is equalizing the room to remove feedback. This is generally
accomplished by turning up system gain to purposely induce feedback, searching for the center frequency of the
feedback, and then equalizing at that frequency to remove
the feedback. Once this frequency has been cut, system
gain is again increased to induce another feedback point,
and the whole procedure is repeated until the engineer is
satisfied that the significant problem frequencies have
been corrected. The major problem with this approach is
that the feedback can easily get out control, and the engineer ends up dashing back and forth between the mixer
volume controls and the equalizer controls, while everyone in the room plugs their ears and prays it will end
soon. The Ashly DPX-100 can turn this procedure into a
fast, painless job, eliminating loud feedback levels and
the possibility of speaker or ear damage.
Procedure:
1. Set up the DPX-100 limiter controls
as follows:
a. Output level control to -20dB.
b. Input Gain control to 0dB.
c. Threshold control to -30dB.
d. Ratio control to infinity (∞)
e. Attack time to 5mS.
f. Release time to 1 Sec.
g. Limit switch IN
2. Using a 1/3 octave (31 band) or parametric
equalizer, set the EQ controls to a flat setting, and if the
equalizer has an overall volume control, boost it by 10 to
15 dB.
3. Open up several microphone input channels
to a normal operating level, with typical EQ settings, and
turn the console master fader up to a louder than normal
setting. At this point, the system should be well into feedback, but the room volume will be constant due to the
action of the limiter. You can listen to the feedback at
any level you like by simply varying the limiter output
level control, although below a certain monitoring level,
the feedback will stop.
4. Try to determine the feedback frequency, and
then equalize it by adjusting the center frequency, bandwidth, and boost/cut controls of your parametric equalizer. (Note: a graphic equalizer can also be used, although
with less accuracy.) After eliminating the problem frequency, try to further define it by sharpening up the bandwidth, reattacking the frequency control, and making the
cut shallower, if possible.
5. As soon as the first feedback frequency has
been removed, the compressor/limiter will automatically
bring up system gain until another feedback point is induced. Repeat the equalization procedure until it becomes
impossible to distinguish individual, predominant feedback frequencies.
6. Write down EQ marks for safekeeping if necessary, and return all mixer, EQ master gain, and compressor/limiter gain controls to normal operational
settings.
Altering the Texture of Musical Instruments
It would be impossible to mention here all the
ways that compression is used to create new sounds with
familiar instruments. Some typical uses are:
1. Creating a “fatter” kick drum or snare sound.
2. “Thickening” acoustic guitars.
3. Adding punch and sustain to electric bass or
guitar.
In general, use a gentle compression ratio, say
4:1, with a 10 mS attack time, 0.1 Sec. release time, and
a low enough threshold to cause 6 to 10dB of Gain Reduction. Try using this effect to help bring out a lead
vocal or instrumental solo in a cluttered mix. The compressor is also a great corrective tool when working with
singers whose own dynamic control is less than perfect.
A little compression helps to keep their quieter lines from
becoming buried in the mix. Experimentation is highly
recommended.
Voice-Over Compression (“Ducking”)
The compressor/limiter can be used to automatically reduce music to a background level when an announcer is speaking. In this scheme, only the music signal
is actually gain-reduced by the limiter. However, the detector is connected to respond to an announcer’s voice
instead of the music’s peaks. Voice-Over compression
assumes you are already using some sort of mixer to combine the music and mic signals. Use the direct out (send)
of the mic channel to feed the detector input on the CLX
unit. Note: Be sure to use a mono plug for the CLX
detector input. Then use the Threshold and Ratio controls to determine when and by how much the announcer’s
voice affects the music level.
De-Essing
A special type of saturation problem often encountered in recording is the sibilant (Ssss) sound of the
human voice. High frequency, sibilant sounds can reach
very high energy levels, so that a voice that is otherwise
undistorted breaks up on the esses, producing a raspy, undesirable sound. With analog recording to magnetic tape,
high frequencies tend to saturate the tape sooner , and combined with the internal high frequency boost (record preemphasis) on standard tape decks, the need to control
sibilants becomes apparent.
The solution is frequency-dependent limiting,
which is easily accomplished with the DPX-100. By inserting an equalizer into the Detector Patch point and
boosting the equalizer at high frequencies in the vicinity
of the sibilant, the limiter’ s detector circuit becomes mor e
sensitive to this particular range of frequencies, and so
will limit the bothersome sibilants more than other frequencies.
Realize that this technique is very different from
simple equalization. Equalizing a sibilant vocal by cutting high frequencies would result in a loss of important
high frequency information at all times, whereas de-essing
has no effect whatsoever on the signal except at the instant of the sibilant. At that moment, the Ashly limiter
will reduce overall gain. Frequency response is unaffected, and the sibilant is controlled.
While most graphic equalizers look very much
the same, there are several important differences in the
circuitry used to implement various designs.
Perhaps the major differences are in the filters.
Some equalizers use a filter made of a capacitor, an inductor, and a resistor, or “RLC” filter. The advantage
here is simplicity, but the real disadvantage is the inductor itself. An inductor is a coil of wire with a core of
some sort. Inductors are susceptible to hum fields and
they are large and expensive.
Other equalizers use the same basic approach, but
replace the inductor with a “simulated inductor”, which
is actually a circuit comprised of an amplifier, a capacitor, and a couple of resistors. This adds parts but is less
expensive than a real inductor. The problem with this
approach is that simulation is less than ideal; it produces
an inductor with high resistive loss resulting in poor curve
shape when used in a filter.
The summing system in a graphic equalizer is
also important. Since there are a number of filters which
combine to produce the overall response, it is important
that the filters not interact (they WILL overlap, but the
response of one filter should not modify the response of
another). Ashly uses an “inter leaved” summing system
where every other filter uses the same summing amplifier
so that adjacent filters never share the same drive and
feedback signals. This allows the filters to maintain their
natural response.
Compressor/Limiters: The Need For Gain Control
The human ear excels in its ability to detect an
extremely wide range of loudness levels, from the quietest whisper to roar of a jumbo jet. When we attempt to
reproduce this dynamic range, by means of amplifiers,
tape recorders, CD players, or radio transmitters, we run
into one of the fundamental limitations of these electronic
media: limited dynamic range. Amplifier dynamic range
is quite good, and is adequate for most musical program
material. However, some types of audio equipment, such
as cassette tape recorders, have a very narrow useful dynamic range.
Another problem with “RLC” designs is that large
capacitors must be used for the lower frequency filters,
limiting the choice to large, expensive non-polar types or
electrolytic capacitors with poor audio performance. Also,
when this filter type is combined with a potentiometer to
adjust the equalization, the resistance of this pot affects
the “Q” of the filter so that a little equalization produces
a much broader curve than a lot of equalization.
The other filter approach is a true bandpass filter. This can be made with no inductors and more practical sized capacitors; the “Q” is easily set and remains
constant, and the parts count is reasonable. there are several types of bandpass filters suitable for this job. Ashly
uses a “Q” enhanced Wein-bridge filter. Because it is a
symmetrical design using matched tuning components,
the “Q” is easily set and is very stable.
In designing a graphic equalizer, a selection of
filter sharpness must be made. More sharpness (higher
Q) produces less filter overlap and tighter control over an
individual band, but also causes ripple in the frequency
response when many filters are boost or cut together to
produce a flat response. We feel that the graphic
equalizer’s primary use is for “voicing” and tone control,
and have set our filter sharpness to produce a maximum
of 1dB ripple.
What is it that compromises the dynamic range
of this equipment? The useful operating region of a piece
of audio equipment is squeezed in between noise and distortion. As program level decreases, it approaches what
is known as the “noise floor”, and if the volume of the
program material goes lower still, it is engulfed by the
noise. The noise floor, or minimum constant noise leve l,
will consist of hiss, hum, transistor noise, tape hiss, buzz
and whatever noises are inherent in the medium. When
the program level is considerably higher than the noise
floor, our hearing masks the noise, and it is not a problem. However, when listening to very quiet sections of a
program for example, a pause between movements of a
string quartet the noise can become very bothersome.
At the other end of the loudness spectrum, the
limitation on dynamic range is usually distortion, either
in the form of amplifier overload, tape satu ration, or A to
D clipping. In most transistorized equipment, the transition from clean, undistorted operation to severe distortion is very abrupt. Therefore, it is common practice to
operate a piece of equipment at a level that is somewhat
below the distortion point, leaving a margin of safety for
unexpected, transient volume peaks in the music. This
safety margin is known as headroom, and may range from
10 to 25 dB. Lowering our standard operating level to
leave ourselves some headroom helps prevent distortion,
but at the same time it moves our ave rage program level
closer to the noise floor, thereby compromising signalto-noise performance. It becomes apparent that to get most
out of an audio system, you have to keep your standard
operating level as high as possible without risking distortion.
GAIN RIDING
One solution to the noise vs. distortion trade-off
is to keep your hand on the level control and manually
adjust gain to suit the program. Indeed, there are times
when this approach is entirely satisfactory. However, in
most types of music there are instantaneous, short duration volume peaks, or transients, which would be difficult
to anticipate and impossible to respond to with manual
gain riding, you simply could not bring the level down
fast enough. In many situations, this can present real problems. For example, in recording, an extra burst of enthusiasm from a lead singer might overload the capabilities
of your recording tape, causing ragged distortion and necessitating another take. In sound reinforcement, a sudden burst of energy through the system can blow fuses or
even damage loudspeakers.
In addition to the problem of response time with
manual gain riding, it also requires your constant attention, which takes you away from more important jobs. The
need for a fast-acting, reliable, automatic gain control is
answered by limiters and compressors.
WHAT COMPRESSORS AND LIMITERS DO
LIMITING
In any musical program are constant changes in
loudness. It is the job of a limiter to detect when the volume has exceeded a predetermined maximum safe level,
and to then turn down the volume. When the incoming
signal returns to its original level, the limiter should respond by restoring the gain to normal. Thus, when the
level is within a specified “safe” range, the limiter has no
effect. When an occasional peak occurs, the limiter responds. This situation is completely analogous to manual
gain riding, except that it occurs faster and more consistently.
COMPRESSION
A very significant difference in dynamic range is
achieved simply by changing the relationship between nominal signal level and threshold, as a result of either increasing the GAIN and/or decreasing the THRESHOLD control.
The most interesting effect to be noted, how ever, is seen by
comparing the original input signal with the output signal.
The quietest portions of the original signal will be effectively increased in volume while the loudest portions of the
original signal will be decreased. In effect, both ends of the
dynamic spectrum will be pushed toward the “middle”.
This is quite different from simple limiting, where only
loud peaks are subjected to gain reduction. More than
anything else, it is this double-ended effect which distinguishes compression from limiting. Compression is
further differentiated from limiting by careful selection
of attack and release times. When limiting is employed
to protect an audio system against transient volume peaks
and possible overload, attack time is usually set as fast
as possible, consistent with distortion-free performance.
Release time would also be relatively short, so that the
output signal would be restored to normal as quickly as
possible after the transient.
Compression is frequently used to keep overall signal level within a specific dynamic range, and
for this application, slower attack and release times
are usually chosen. This approach is analogous to our
manual gain riding example, where our operator is fading the music up and down to keep it fairly constant,
but is doing it slowly enough so that the listener is unaware that the gain is being altered.
Voltage Controlled Amplifiers
Early VCA’s wer e based on vacuum tubes with
a “remote cutoff” characteristic. The tube would simply change its gain in response to a changing bias voltage. Tubes developed for this purpose did an excellent
job, in fact they could exceed the noise and distortion
performance of today’s best solid state VCA’s. Unfortunately, they also had some serious disadvantages peculiar to tubes - change of gain and matching as aging
took place, heat, microphonics, high cost, and the need
for both high-voltage and filament power supplies.
Over the years the need for good, low-cost,
solid state VCA brought about many innovative approaches. A good example is the electro-optical attenuator where a photocell is used as one leg of a
potentiometer. Since the photocell behaves as a true
resistor, distortion and noise are very low. Unfortunately, the response time of photocells is slow and unpredictable so their use in a fast peak-limiter is really
not feasible. Also, the matching between units is very
poor so that stereo tracking is not possible without tedious hand-matching of photocells.
Another approach uses a field-effect transistor (FET) as a variable resistor. Here, at least, the response time is fast (in the nanosecond range), but
matching between units is still poor, requiring hand
matching for stereo. An additional problem is that a
FET will only act as a pure resistor with very small
signals applied so it is necessary to attenuate an input
signal before the gain control FET and then amplify it
again. Of course this results in less than ideal noise performance and imposes a frustrating trade-off: less noise
= more distortion.
A number of VCA’ s based on the exponential voltage-current characteristic of a bipolar junction transistor
have been used. One of the most common is called a
“transconductance amplifier”. Using the inherent matching obtained by integrated circuit technology, these devices have very predictable control characteristics.
Tracking within 1dB over a 40dB range is common. Not
only do the control characteristics match well from unit
to unit, but they can easily be made exponential (logarithmic) so that even increments of control voltage produce even increments of gain change in decibels. The
response time is also very fast.
The problem with simple transconductance amplifiers is that, like FET VCA’s, they can handle only very
small signals so the noise performance is poor. A number of linearizing circuits have been devised to minimize
this problem, but even the best transconductance amplifiers have an equivalent input noise of about -80dBv, which
compares poorly to straight linear amplifiers.
The best analog compromise to date is the “class
AB current ratio multiplier.” Early implementation of
this circuit used two matched pairs of transistors, one pair
of NPN’s and one pair of PNP’s. The problem here is
that excellent matched integrated NPN pairs were available, but integrated PNP’s were not. The PNP’s had to be
hand-tested and matched. Careful trimming was necessary for low distortion and even minor temperature
changes made re-trimming necessary because of differing characteristics between the two types.
The Ashly VCA
The Ashly VCA is an inte grated current ratio multiplier circuit. It has low noise (-90dBv), low distortion
(.05%), excellent response time and tracking and does
not suffer from thermal drift. The noise and distortion
are at state-of-the-art levels and the circuit is consistent
in mass production with minimal trimming and no handselection of transistors.
Detectors
It would seem that, of the two components in a
compressor/limiter, the VCA is the more critical since the
audio passes through it and the detector only provides it
with a control voltage. Experience showed us that both are
crucial to the overall sound and that, if anything, the
detector’s perf ormance is the harder to judge by conven-
tional measuring techniques. While the VCA is doing its
job if it has low noise and distortion, the detector must constantly adjust the gain of the audio path in a manner which
keeps the level under control while sounding acceptable to
the listener. This constantl y changing gain is a dynamic
action, while conventional audio measurements like noise
and distortion checks are Static (at a constant level). We
became painfully aware of this problem with some of our
earlier limiter prototypes which measured fine and sounded
terrible. This led us to use a purely subjective approach in
the design of the detector - we did a lot of listening to determine what sounded good and what didn’t.
Two important features emerged from this re-
search:
1. We designed the detector to let the attack and
release times speed up as more and more limiting occurs.
The compression ratio also increases. This lets us maintain peaks fairly close to a constant ceiling level, but allows the illusion of increasing loudness as input level
increases, thereby preventing complete loss of dynamics
when limiting.
2. We incorporate a double release time constant.
When release time was set slow with a single time constant, transients such as mic “pops” caused a quick reduction in gain and a slow fade-up, making the action of
the limiter very obvious. With the double time constant,
release from gain reduction after a brief transient is always fast, with a slower release after a sustained overdrive.
When choosing a compressor/limiter, you can see
that it is very important to listen to it in your particular
application and see that it sounds the way you want. There
are lots of these devices with seemingly excellent specs
which sound very different with real program material
applied to them.
Peak Or RMS
There are several ways of looking at a signal to
determine its level. A peak detector looks at the maximum voltage a signal reaches regardless of it’s waveform,
while an RMS (root mean square) detector looks at the
energy in a signal regardless of the short term voltage
levels. This makes a peak detector the correct choice for
preventing clipping, overmodulation, or tape saturation,
while an RMS detector can be used to restrict material to
a given loudness. When an RMS limiter is used to pre vent clipping, the result is unpredictable. For instance, a
flute and a snare drum which are limited to the same RMS
level might have peak levels as much as 30dB apart! Use
peak limiters to prevent clipping.
Check AC power - is the pilot light on?
Check in/out connections - are they reversed?
Are you sure you have an input signal?
6dB/15dB
15 BAND EQ
EQ In/OutHPF In/Out
VCA
CHANNEL
IN/OUT
TH
THRESHOLD
OUTPUT
LEVEL
Distorted Sound
This will only be caused by too much
signal which will show on the Clip LED. If the
LED is not flashing, there is an overload some-
SIG
where else in the signal path. Adjust the relative gain of each component in your chain to
CLIP
2
3
1
keep everything at a comfortable level.
Excessive Hum or Noise
Hum will usually be caused by a ground
loop between components. Try using the sug-
EQ OUTPUT
gested balanced input and output hook-ups if the
other pieces of equipment used in conjunction
with your equalizer have balanced inputs and
outputs.
CLIP
Noise (excessive hiss) can be caused by
insufficient dri ve signal. Make sure you are send-
2
3
1
ing a nominal 0 dBu line level signal to the
equalizer. Most noise problems occur because
gain is applied to audio signals too late in the
chain. For best performance, apply gain to in-
LIMITER
LIMITER
dividual source signals as early as possible, like
at the mixer input preamp section. As gain increases, it also boosts the noise content of that
DETECTOR
INSERT
signal. An y cumulative noise b uilt up in a mixe d
signal will only be increased by using an equalizer as a gain device, so make eve ry attempt to
operate the equalizer with as little gain as possible.
Note: Unshielded cables, improperly wired connectors, and cables with broken strands of wire are very
common problems. Use quality cables with quality, correctly wired connectors.
11.1 Compressor/Limiter Troubleshooting Tips
EQ Controls Do Nothing
Is the master EQ switch in?
The lowest and highest frequency sliders may be
beyond the range of the program material or speakers and
may produce little or no audible effect.
Peak Light Flashes or Stays On All the Time
If the peak light flashes, the signal level to the
equalizer is too high. Turn down the gain. If it is on all
the time, disconnect the input and output cables. If it is
still on, the unit must be returned for service.
14
No Output
Check A C power . Is the po wer switch on? Check
input and output connections - are they reversed? Are
you sure you have an input signal?
Controls Have No Effect
Is the limiter in/out switch in? Perhaps the ratio
control is set too low to produce an audible effect or the
input lev el is belo w threshold. Is the threshold LED lighting up? If not, lower the threshold setting or increase the
gain. Do not expect to hear any effect when the input
level is below threshold, since the unit is simply a linear
amplifier at those levels.
When Using Heavy Compression, Background Noise
Is Noticeable During Quiet Sections Of The Program
As defined in the section on compression, quiet
program material is effectively made louder while loud
peaks are made quieter. When the program source is thus
raised in volume, its noise floor is also raised in volume
by a proportionate amount. This is not a defect in the
compressor/limiter, but an unavoidable side effect of the
gain altering process. If the noise becomes a problem,
the solutions are to either decrease noise at the program
source, or use less compression.
Excessive Hum Or Noise
Hum is often caused by a “ground loop” between
components. Try using the suggested balanced input and
output hookups if the other pieces of equipment used in
conjunction with the DPX-100 have balanced inputs and
outputs. Noise can also be caused by insufficient drive
levels. Make sure you are sending a nominal 0 dBV line
level signal to the unit.
12. WARRANTY INFORMATION
Thank you for your expression of confidence in
Ashly products. The unit you have just purchased is protected by a five-year warranty. To establish the warranty,
be sure to fill out and mail the warranty card attached to
your product. Fill out the information below for your
records.