Antares AVOX 2 user manual

							25
Chapter 8: SYBIL
Variable Frequency De-Esser
SYBIL is a digital 
equivalent of a 
traditional vocal   
de-esser.
When recording spoken 
or sung material, the 
sibilant sounds (Ss, Ts, 
CHs, and SHs) in the track 
sometimes (depending on 
the individual performer) 
appear louder than the 
rest of the signal. The 
effect can sound unnatural 
and often irritating. The 
solution to this problem 
is to compress only the 
sibilants, thereby lowering 
their level relative to the rest of the track. Processing a signal this way is called de-essing. 
The diagram below shows how a compressor and a high pass filter are traditionally configured to 
accomplish de-essing.
SYBIL uses a digital algorithm to implement 
the de-esser function. While the details of 
the algorithm are quite complex, the resulting 
effect is functionally equivalent to the diagram 
above. 
Controls
Gain Reduction Meter
The Gain Reduction Meter displays the amount 
of compression taking place. If the frequency 
and threshold controls are set properly, the 
meter will display little gain reduction during 
vowel sounds and soft consonants, and 
substantial gain reduction during sibilants. For 
this reason, the meter is a useful tool when 
setting the other controls.
High Pass Frequency
This control sets the frequency of SYBIL’s 
side-chain high pass filter. The goal is to set 
the frequency such that the filter passes any 
sibilance (thereby keying the compressor), but 
not any of the desired signal.
COMPRESSOR
IN IN
OUT
OUT
SIDECHAIN 
INPUT
HIGH
PASS
FILT ER 
						

							26
Command (Mac)/Control (PC) click the control 
to reset it to its default value of 8000 Hz.
Threshold
This control sets the threshold level of SYBIL’s 
compressor. The range is from 0 dB to - 80 dB. 
In most cases, the default value of -48 dB is a 
good place to start.
Command (Mac)/Control (PC) click the control 
to reset it to its default value of - 48 dB.
Compression
This control sets the compression ratio of 
SYBIL’s compressor. The range is from 1.0:1 to 
99:1. In most cases, the default value of 2.4:1 is 
a good place to start.
Command (Mac)/Control (PC) click the control 
to reset it to its default value of 2.4:1.
Attack Time 
This control adjusts the speed with which 
SYBIL’s compressor responds to peaks in the 
signal coming from the high pass filter. The 
range of the control is from 3 milliseconds to 
100 milliseconds. 
Command (Mac)/Control (PC) click the control 
to reset it to its default value of 10 milliseconds.
Release Time
This control adjusts the time it takes the 
compressor’s gain to increase 6 dB after the 
signal coming from the high pass filter drops 
below the threshold. The range of the control is 
from 3 milliseconds to 200 milliseconds.
Command (Mac)/Control (PC) click the control 
to reset it to its default value of 20 milliseconds.
Usage Tips
• SYBIL functions solely as a mono processor.
• If the high pass frequency is set too low, 
non-sibilant components of the signal will 
be compressed and the vocal will have its 
highs attenuated. If it is set too high, some 
sibilance will still remain. The trick is to find 
that ideal point where only the sibilance is 
affected. When set correctly, you can’t tell 
that there is processing going on. If you can 
hear something happening, SYBIL needs to 
have its settings tweaked. 
						

							27
Chapter 9: ARTICULATOR 
Vocal Formant and Amplitude Modeler
The alien offspring of a vocoder 
and a modern-day version of the 
venerable talk box, ARTICULATOR 
lets you extract the formant and 
amplitude information from a vocal 
(or other dynamic source) and apply 
it to any other audio track or to a 
built-in broadband noise generator.
Perfect for talking guitars, singing synths, 
whispering voices, and a wide range of  
special effects.
So what’s going on here?
ARTICUL ATOR is unique among the AVOX 2 
plug-ins in a number of basic ways:
First, it requires two different audio sources, 
what we call the Audio Signal and the Control 
Signal (more about those in a bit). 
Second, while every other AVOX 2 plug-in is 
designed to process and affect a vocal track,  ARTICULATOR is designed to use a vocal track 
to control the processing of a second track, 
ideally a track with broadband harmonic content 
like a rich synth patch, a processed guitar, or 
the built-in noise generator (but typically not 
another vocal track).
Finally, because different hosts offer different 
(or no) capabilities for routing multiple tracks to 
a single plug-in, the version of ARTICULATOR 
for each plug-in format features a unique Input 
Routing subsection that is designed for the 
specific capabilities of that format.
Signal Selection
As mentioned above, ARTICULATOR requires 
two audio sources (although the built-in Noise 
Generator can serve as one of them). They are:
The Control Signal: This signal serves as 
the source of the formant and amplitude 
information that will be applied to the Audio 
Signal. A good Control Signal will have a lot 
of variation in both loudness and resonant 
harmonic content. A voice, whether singing, 
speaking, or just making weird rhythmic vocal 
noises makes an ideal Control Signal.
The Control Signal will always be mono. 
						

							28
The Audio Signal: This is the signal that 
the Control Signal’s formant and amplitude 
information is applied to. A good Audio Signal 
will be rich in harmonic content and sustained 
in nature. Examples include synth pads and 
patches rich in overtones, distortion-rich guitars 
sounds, or even entire instrumental mixes. The 
built-in Noise Generator also makes a good 
Audio Signal, but its character will pretty much 
always be in the nature of whispering. (It can 
also be mixed with an external Audio Signal.) 
The Audio Signal can be mono or 
stereo, depending on the capabilities 
of your particular host.
Signal Routing
Since few current plug-ins need to 
make use of two independent audio 
signals, some ingenuity is required to 
use ARTICULATOR in some hosts. 
The easiest way to route the required signals 
is through the use of a sidechain input. In this 
scenario, you simply instantiate ARTICULATOR 
on the Audio Signal’s track and route the 
Control Signal through the sidechain input.
Unfortunately, not all hosts currently support 
sidechain routing. If yours doesn’t, you will 
have to use one of the alternative methods 
outlined in the Input Routing section below.
Controls
Input Routing Section
As mentioned above, the RTAS, VST and 
Audio Units versions of ARTICULATOR each 
include an Input Routing section unique to each 
format’s capabilities. In the case of RTAS and 
VST, these sections are simply informational in 
nature, providing an indication of what’s going 
where. In the case of Audio Units, the Input 
Routing section includes controls that let you 
select the routing that is supported by your 
particular host. Details below. R TA S
 Since Pro Tools
™ 
supports sidechain routing, 
setting up the RTAS version 
is quite straightforward. 
Simply select the Audio 
Signal by instantiating 
ARTICUL ATOR on the 
desired mono or stereo 
audio track and then assign 
the Control Signal via the 
sidechain input. 
The RTAS Input Routing 
section includes a red 
“LED” that lights to confirm 
that the sidechain input has 
been enabled in Pro Tools.
VST  The version of VST 
that is most commonly 
in use at the time this is 
being written (and is used 
by ARTICUL ATOR) does 
not support sidechain 
input (but see the note 
below). Consequently, the 
methods for getting both 
signals into the VST version 
of ARTICULATOR vary 
depending on what kind of 
track it’s instantiated on. 
The Input Routing section 
will change to reflect each 
of the following three 
possible options:
Instantiation on a mono 
track: In this configuration, 
instantiate ARTICULATOR 
on the track of the Control 
Signal. ARTICULATOR’s 
internal Noise Generator  
						

							29
functions as the Audio 
Signal. (This is useful 
primarily for whispering 
and modulated noise 
effects.)
Instantiation on a stereo 
track: In this configuration, 
ARTICUL ATOR treats the 
right channel of the stereo 
track as the Control Signal 
and the left channel of the 
track as the Audio Signal. 
Keep in mind that this 
will usually require that 
you create the stereo 
track by combining (via 
whatever method your 
DAW provides) the two 
independent tracks you 
wish to use as Control 
Signal and Audio Signal 
(being careful to ensure 
that they are hard panned 
to the appropriate 
channels). 
In this configuration, the 
Audio Signal is limited to 
being mono (although it 
can be combined with the 
Noise Generator in Stereo 
Noise mode).
Instantiation on an 
LSR Group Channel: 
Instantiation on an 
LRS Group Channel: 
In this configuration, 
ARTICUL ATOR treats 
the surround channel of 
the group as the Control 
Signal and the left and 
right stereo channels 
as the Audio Signal. 
Again, this will require 
that you create the 
LRS Channel signals by 
combining the independent 
tracks you wish to use as 
Control Signal and Audio  Signal (again being careful to ensure that they 
are hard panned to the appropriate channels).
This is the most flexible of the three 
options, in that it allows you to 
process a stereo Audio Signal.
N OTE: A new version of VST that 
does include sidechain support 
was released just as AVOX 2 was 
being released. It’s not yet clear how quickly 
this version will be supported in all hosts. 
There is the possibility (but no guarantee) that 
a future version of ARTICULATOR may add 
sidechain support for VST. If so, it would be 
noted in an accompanying Read Me file. 
AUDIO U NITS
Some Audio Units 
hosts support sidechain 
input and some don’t. 
Consequently, the Input 
Routing section for the 
AU version provides 
controls that allow you to 
set up ARTICUL ATOR 
to take advantage of 
whichever capabilities 
your host provides.
The Input Routing section 
will change to reflect 
each of the two possible 
options:
Instantiation on a mono 
track: In this configuration, 
you can select either the 
sidechain input or the 
audio track itself as the 
Control Signal. If your host 
supports sidechain input, 
you will typically always 
select it as the Control Signal. If your host does 
not offer sidechain input, select Audio Track. 
Depending on your choice, the resulting Audio 
Signal source will appear in the Audio Signal 
display as follows:
•  If you select Sidechain, the Audio Source 
will be the audio track that ARTICULATOR is 
instantiated on. 
						

							30
• If you select Audio Track, ARTICULATOR’s 
internal Noise Generator functions as the 
Audio Signal. (This is useful primarily for 
whispering and modulated noise effects.)Input Mixer Section
The Input Mixer controls let you set the relative 
levels of the Audio Signal and the internal 
Noise Generator as well as optionally applying 
compression to the Audio Signal and selecting 
mono or stereo noise.
AUDIO LE VEL
This control sets the 
level of the audio 
signal (duh!). Its 
range provides for 
both attenuation and 
gain. The range is 
from -inf (i.e., off) to 
+12 dBs of gain.
To adjust Audio Signal Level, click on the 
knob and move the cursor up to turn the knob 
clockwise, or down to turn it counterclockwise. 
The numeric display will indicate the amount of 
attenuation or gain in dBs.
Command (Mac)/Alt (PC) clicking on the knob 
will reset it to its default value of 0 dB. COMPRESSOR
This control lets you engage a 
compressor with fixed parameters 
designed to maximize the Audio 
Signal’s level. It would typically be 
used in the case where the Audio Signal has 
quite a bit of dynamic variation that you would 
like to even out so that the Control Signal’s 
dynamics can be applied to it. 
Clicking the Comp button will toggle the 
compressor on and off. The button will turn 
blue to indicate that the compressor is on. 
N OTE : Engaging the compressor 
usually results in an overall increase 
in gain. You may need to adjust the 
Audio Level control to compensate.
NOISE LE VEL
This control sets the 
level of the internal 
Noise Generator. 
The range is from 
-inf (i.e., off) to +12 
dBs of gain.
Instantiation on a 
stereo track:
 In this 
configuration, you can 
select either the sidechain 
input or the left channel of 
the audio track or the right 
channel of the audio track 
as the Control Signal. 
If your host supports 
sidechain input, you will 
typically always select it 
as the Control Signal. If 
your host does not offer 
sidechain input, select 
either the left or right 
channel of the audio track. 
Depending on your 
choice, the resulting 
Audio Signal source will 
appear in the Audio Signal 
display as follows:
•  If you select Sidechain, the Audio 
Source will be the stereo audio track that 
ARTICULATOR is instantiated on. This is the 
most flexible configuration in that it allows 
for a stereo Audio Signal.
•  If you select Audio Track L as the Control 
Signal, Audio Track R will be selected as the 
Audio Signal.
•  If you select Audio Track R as the Control 
Signal, Audio Track L will be selected as the 
Audio Signal.
In these latter two cases, you will typically be 
required to create the stereo track by combining 
(via whatever method your DAW provides) the 
two independent tracks you wish to use as 
Control Signal and Audio Signal (being careful 
to ensure that they are hard panned to the 
appropriate channels). In these configurations, 
the Audio Signal is limited to being mono 
(although it can be combined with the Noise 
Generator in Stereo Noise mode). 
						

							31
To adjust Noise Generator Level, click on the 
knob and move the cursor up to turn the knob 
clockwise, or down to turn it counterclockwise. 
The numeric display will indicate the 
amount of attenuation or gain in dBs.
Command (Mac)/Alt (PC) clicking on the knob 
will reset it to its default value of -inf (i.e., off).STEREO NOISE
ARTICUL ATOR’s internal 
Noise Generator can provide 
either a single noise signal, 
or two uncorrelated noise 
signals panned to the 
opposite extremes of a stereo output. 
Clicking the Stereo Noise button will 
toggle it between mono and stereo 
noise. The button will turn blue to 
indicate the selection of stereo noise. 
N OTE : Stereo noise is only 
available from stereo or 
mono->stereo versions of 
ARTICULATOR. If you have instantiated a 
mono version, this button will not appear.
Formant Extraction Section
The controls in this section allow you to 
optimize the formant extraction process for 
your specific Control Signal.
VOICE T YPE
Use this control 
to select the 
range of the 
track you will 
be processing. Choices include soprano voice, 
alto/tenor voice, baritone/bass voice and 
instrument (a general setting for anything that 
isn’t actually a vocal).
Matching the appropriate algorithm to the input 
results in more accurate formant extraction.
To select voice type, click on the Voice Type 
pop -up and then select the desired range from 
the pop -up list.
N OTE : Choosing the wrong Voice 
Type (or just forgetting to set it 
at all) can result in compromised 
performance. Pay attention. TR ACKING
In order to 
accurately analyze 
the Control Signal, 
ARTICULATOR 
requires a 
periodically 
repeating 
waveform, 
characteristic of a voice or solo instrument. 
The Tracking control determines how much 
variation is allowed in the incoming waveform 
for ARTICUL ATOR to still consider it periodic.
If you are working with a well-isolated solo 
Control Signal (e.g., tracking in a studio or from 
a multitrack master) you can typically set the 
Tracking control to 25 and forget it.
If, on the other hand, your Control Signal is 
noisy or not well-isolated, or you are dealing 
with a particularly breathy or guttural voice, 
it may be necessary to allow more signal 
variation (higher Tracking numbers). However, 
under these circumstances tracking is not 
guaranteed and a too “relaxed” setting may 
negatively affect the accuracy of the formant 
extraction.
To adjust Tracking, click on the knob and move 
the cursor up to turn the knob clockwise, or 
down to turn it counterclockwise. Remember, 
higher numbers allow more  variation, while 
lower numbers require less.
Command (Mac)/Alt (PC) clicking on the knob 
will reset it to its default value of 25. 
						

							32
Formant Modulation Section
Another way of looking at the Control Signal’s 
formant information is to think of it as a real 
time model of the vocal tract of the person 
singing or speaking. The controls in this section 
allow you to modify the extracted formant 
information by modifying the geometry of that 
modeled throat.THROAT LENGTH
This control lets 
you increase or 
decrease the 
length of the 
modeled throat 
that is applied to 
the Audio Signal.
Values above 
1.00 represent a 
lengthening of the throat while values below 
1.00 represent a shortening of the throat. The 
actual values represent the percentage change 
in the throat length. For example, a value 
of 1.20 represents a 20% increase in throat 
length, while a value of 0.80 represents a 20% 
decrease in throat length.
The effect of this control is so dependent on 
the combination of each specific Control Signal 
and Audio Signal that the only guideline for 
working with it is to just check out its entire 
range and discover what works best for your 
particular needs.
To adjust Throat Length, click on the knob and 
move the cursor up to turn the knob clockwise, 
or down to turn it counterclockwise. The 
numeric display will indicate the percentage of 
change.
Command (Mac)/Alt (PC) clicking on the knob 
will reset it to its default value of 1.00. THROAT WIDTH 
As you might 
imagine, this control 
is similar the the 
Throat Length 
control described 
above, but instead 
modifies the width 
of the modeled 
throat.
Values above 1.00 represent a widening of 
the throat while values below 1.00 represent 
a narrowing of the throat. The actual values 
represent the percentage change in the throat 
width. For example, a value of 1.20 represents 
a 20% increase in throat width, while a value 
of 0.80 represents a 20% decrease in throat 
width.
Again, your best bet in working with Throat 
Width is just to try a variety of settings to hear 
what works best for your project.
To adjust Throat Width, click on the knob and 
move the cursor up to turn the knob clockwise, 
or down to turn it counterclockwise. The 
numeric display will indicate the percentage of 
change.
Command (Mac)/Alt (PC) clicking on the knob 
will reset it to its default value of 1.00.
Amplitude Extraction Section
In the schematic signal flow, this section 
represents the process of extracting the 
Control Signal’s amplitude envelope (which is 
a fancy way of saying “the way its loudness 
changes over time”). 
ARTICULATOR does this all by itself with no 
need for any input from you.  
						

							33
Amplitude Modulation Section
Here’s where you decide how much effect the 
Control Signal’s amplitude envelope will have 
on the Audio Signal.ENVELOPE AMOUNT
This control 
sets the amount 
of amplitude 
modulation 
(i.e., change in 
loudness) applied 
to the Audio Signal 
by the Control 
Signal’s amplitude 
envelope.
When this control 
is set at 0, the Control Signal’s amplitude 
envelope has no effect on the Audio Signal. In 
this case, the loudness dynamics of the Audio 
Signal are preserved and only the formant 
information from the Control Signal is imposed 
on the Audio Signal.
When this control is set at 100 (maximum), the 
the full effect of the Control Signal’s amplitude 
envelope is applied to the Audio Signal. 
And as you might imagine, intermediate 
settings of this control apply intermediate 
amounts of amplitude modulation.
N OTE : In order for the Control 
Signal’s amplitude envelope to have 
maximum effect on the Audio Signal, 
the Audio Signal should be a sustained sound 
(like the Noise Generator or a synth pad) or at  
least one in which audio is constantly present. 
If there are periods of silence in the Audio 
Signal, its output during those periods will 
be silence, regardless of the Control Signal’s 
envelope. If the Audio Signal is constant, 
but has a varying dynamic range, you might, 
depending on your goal, consider engaging the 
Compressor in the Input Mixer to even it out.
TI p: The easiest way to 
understand the effect of this 
control is to do the following:
  1.  Do whatever your DAW requires to set 
up ARTICULATOR with a speaking or 
singing voice as the Control Signal.
  2.  In the Input Mixer, set Audio Level to -inf 
and Noise Level to a comfortable level.
  3.  Set the Formant Extraction controls as 
appropriate for your Control Signal.
  4.  In the Output Mixer, set Control 
Signal Level to -inf and Audio Signal 
level to a comfortable level.
  5.  While playing the Control Track, vary 
the Envelope Amount control from 
0 to 100 and listen to the result.
To adjust the Envelope Amount, click on the 
knob and move the cursor up to turn the knob 
clockwise, or down to turn it counterclockwise.  
Command (Mac)/Alt (PC) clicking on the 
knob will reset it to its default value of 100 
(maximum amount).
EQ Section
The EQ Section provides one band of traditional 
parametric EQ for fine-tuning the Audio Signal 
(as well as the Noise Generator output, if used).  FREQUENCY
This control selects the 
frequency of the EQ (i.e., the 
frequency that will either be 
increased or decreased via 
the Gain control). The range is 
from 100Hz to 4000Hz. 
To adjust Frequency, click 
on the knob and move the cursor up to 
turn the knob clockwise, or down to turn it 
counterclockwise.  
Command (Mac)/Alt (PC) clicking on the knob 
will reset it to its default value of 1500Hz. 
						

							34
Q
The Q control allows you to set 
the slope (and, consequently, 
the bandwidth) of the EQ’s peak 
or dip. Its range is 0.10 to 3.10. 
High Qs result steep slopes (and 
narrow bandwidth), while lower 
Qs result in more gentle slopes 
(and wider bandwidth). (The 
easiest way to understand this is to watch the 
green EQ plot while adjusting the Q and Gain 
controls. All will become clear.)
To adjust Q, click on the knob and move the 
cursor up to turn the knob clockwise, or down 
to turn it counterclockwise.  
Command (Mac)/Alt (PC) clicking on the knob 
will reset it to its default value of 0.71.
GAIN
The Gain control lets you set the 
amount of boost or cut that will 
be applied to the Audio Signal 
at the selected Frequency. 
The range is from -12.0dB to 
+12 . 0 d B .
To adjust Gain, click on the 
knob and move the cursor up to turn the knob 
clockwise, or down to turn it counterclockwise.  
Command (Mac)/Alt (PC) clicking on the knob 
will reset it to its default value of 0.0.
BYPASS
The Bypass button 
allows you to bypass 
the effect of the EQ section. When the button 
is red, EQ is bypassed. Click the button to 
toggle its state. DISPL AY
The EQ Section Display provides a real-time 
graphic representation of the Control Signal (red 
plot), the Audio Signal (blue plot) and the EQ 
curve (green plot). 
Output Mixer Section
Here’s where you can set the output levels and 
pans of both the processed Audio Signal and 
the Control Signal.
AUDIO SIGNAL LE VEL
This control applies 
gain or attenuation 
to the processed 
Audio Signal plus 
Noise (if any).  The 
range is from -inf 
(i.e., off) to 12 dBs 
of gain.
To adjust Audio Signal Level, click on the 
knob and move the cursor up to turn the knob 
clockwise, or down to turn it counterclockwise. 
The numeric display will indicate the amount of 
attenuation or gain in dBs.
Command (Mac)/Alt (PC) clicking on the knob 
will reset it to its default value of 0 dB. AUDIO SIGNAL 
PA N
This slider allows 
you to set the 
Audio Signal’s 
location in the 
stereo spectrum.