XTA 5 Series - Manual (Page 53)

✁ ✂ ✄ ☎ ✂ ✆ ✁ ✂ ✄ ☎ ✂ ✆ ✁ ✂ ✄ ☎ ✂ ✆ ✁ ✂ ✄ ☎ ✂ ✆ Operator’s Manual Page 53
“D-Max” Clip Limiter
The main limitation with traditional dynamics control is the inability of the processing to react truly instantaneously to the
signal. One of the most significant advantages of digital signal processing over analogue is the ability to delay the audio
signal precisely and without extensive complex hardware. The entire domain of digital signal processing is based around
the combination of delaying, multiplying, and accumulating numbers (representing samples of audio) to implement all the
filters and dynamics processing we have come to expect today.
In the case of dynamics processing, being able to delay a signal allows the processor module to delay the main signal in
relation to the sidechain (the signal being monitored relative to the threshold), so that it can compensate for peaks prior to
the arrival of the main signal.
Consider the situation of a monitor engineer listening to a band perform. Having no access to dynamics processors, he has
had to resort to manually ‘riding the faders’ in an attempt to keep control of the levels. Should the level of one of the
channels on his desk reach an unacceptably high level, he will turn it down appropriately.
There is a hidden sidechain in operation even in
this case. The main signal path is fed through the
monitor desk and the gain controlled by adjusting
the fader. The sidechain is formed by the
feedback path between the engineer’s ears
checking the level and his brain instructing his
hand to turn the fader down if the volume goes
over the threshold he has chosen.
In this case, the delay between the signal actually going over the threshold, the engineer registering the situation, and then
turning the signal down will be in the order of several hundred milliseconds at best. This will only be true if he is not
distracted – in reality, it may be several seconds before any gain reduction is imposed on the signal to bring it under
control.
For an analogue dynamics processor, the situation is much better.
Controlling the gain electronically, and not relying on a human
sidechain feedback mechanism, it can react much more quickly.
The red waveform represents the input to the dynamics module,
with the dotted line showing the threshold for gain control to
occur. There are several peaks towards the start of this signal that
are above the threshold, and so the dynamics processing should
react to these as appropriate. (In this case reduce the gain).
The blue waveform shows the output of the dynamics module. The
circled peak demonstrates that the processor has missed the first
peak above the threshold (as it is very fast and short), but has
‘caught up’ shortly afterwards, keeping all other peaks under
control. As it is unable to predict what is coming, this will always be
a failing with analogue dynamics processing.