TC Electronic SYSTEM 6000 MKII Algorithm Manual - Backdrop - Tutorial
107
bACkDrOP - TuTOrIAl
controls and to listen to the amount of noise reduction,
the occurrence of musical noise, and the smearing of
transients (signal distortion). These variables must be
balanced when using BackDrop to remove noise from
your audio. Fortunately, BackDrop provides you with a
comprehensive set of tools to make the optimal possible
trade-off.
When using BackDrop’s controls it is important to
remember that Aggression and Max Reduction work
together and so you should go back and forth between
these parameters in order to find the best settings. This
also applies to the Basilar Dispersion and Transient
Recovery controls which work together.
Advanced Use
The tutorial in the previous section was intended to give
you a feel for some of BackDrop’s controls. In this section
we endeavor to provide you with some guidelines for
obtaining the best possible performance from BackDrop.
The following steps have been found to be a suitable
approach for getting good performance under almost
all circumstances. As you become more familiar with
BackDrop you will no doubt develop your own techniques
and strategies that work best for you.
As always, the first step in the noise reduction process
is to capture a sample of the audio for the noise print
information. It is very important that the audio used to
calculate the noise print information contain only the noise
and none of the signal.
With the noise print information calculated, use the
Algorithm Model control to select the BackDrop algorithm
(Basic, MultiBand or MultiType) that is best suited to the
noise you are trying to remove.
BackDrop’s fundamental philosophy is to target the noise
reduction to where it is needed most and to apply as little
processing as possible to the signal. The strategy outlined
in this section is in keeping with this philosophy. Generally
speaking, to get the best results from BackDrop, do not
be “over aggressive”. Try to use as low an amount of
aggression, and Max Reduction as possible, and use a
moderate amount of Basilar Dispersion (40% to 60%).
It is useful to begin by getting an idea of how much noise
reduction is necessary for the audio sequence that you
are working on. In many cases reducing the noise by only
a few decibels will make a dramatic improvement in the
quality of the audio sequence. In most cases, you will
not need more than about 10 dB of noise reduction. To
determine how much noise reduction is necessary, set
Basilar Dispersion and Transient Recovery to 0%. Set
Aggression to 100% and Max Reduction to 0 dB. Slowly
increase Max Reduction and listen as the level of noise
goes down. Use the Bypass button to turn the processing
on and off. Adjust Max Reduction until you obtain the
appropriate amount of noise reduction.
With Aggression set to 100%, you are probably applying
more processing than necessary for the amount of noise
reduction that you desire. Therefore, slowly lower the
amount of Aggression until you here the level of the
noise floor begin to increase. This level of Aggression
corresponds roughly to the amount of reduction that you
desire and the value of Max Reduction that you have
selected. Note that this relation between Aggression and
Max Reduction depends on the signal being processed and
so the relative settings will be different for different audio
sequences.
It is now a good time to monitor BackDrop’s output very
carefully and decide the following:
Is this too much noise reduction? Often it is desirable
to leave some background noise in the signal so that it
doesn’t sound unnaturally clean. This is primarily true for
an overall mixed signal rather than for individual tracks that
are yet to be mixed together.
Do you hear any artifacts? If you hear artifacts then
increase the amount of Basilar Dispersion until the level
of the artifacts becomes acceptable or the artifacts
becomes completely inaudible. In some cases, increasing
the amount of Aggression slightly when applying Basilar
Dispersion can also reduce the artifacts. You can also
take advantage of some of the residual hiss to mask
artifacts created by the process itself. By using a little less
noise reduction and aggression, leaving a small amount
of residual hiss, you will have a harder time hearing any
artifacts.
Are the transients (attacks) being smeared? If increasing
the amount of Basilar Dispersion has caused a smearing
of the transients then you may want to increase the level of
Transient Recovery. It should be noted that higher levels of
Transient Recovery can cause some of the musical noise
to reappear. Therefore, Basilar Dispersion and Transient
Recovery need to be adjusted together in order to minimize
both the musical noise and the smearing of transients.
Are you losing the sound quality or clarity of the original
source? It is easy to become very focussed on the task of
removing noise and overlook the fact that the signal has
become somewhat distorted (altered). Remember to listen
to the signal as well as the noise.You will need to strike a
balance between the amount of noise reduction possible
and the effect on the sound quality of the music. Keep in
mind that BackDrop was specifically designed to enable
you to make the best possible trade-off for your application.
You should be aware that when removing noise it is
common to uncover low-level sounds that you might think
are artifacts resulting from BackDrop. Frequently these
sounds are actually part of the original signal but they were
partially hidden by the noise. Use the Bypass button to
compare the processed signal to the original signal to see
if the sound was actually part of the original.
While keeping the above points in mind, you should go
back and forth between the various controls and adjust
them until the optimal noise reduction solution is reached.
Of course, you should try all three of BackDrop’s algorithm
models to determine which one is best suited for the task
at hand. Remember that you can switch instantly between
the algorithm models and the settings for each model are