User Manual

17
About level and resolution in the L2
Maximum level
The maximum level of a digital signal is generally governed by the source’s
highest peak. Today’s demanding production environments require the
average levels to be much higher than in the past, and with the accom
-
panying risk of digital clipping comes the need for precision control of the
maximum peak level.
The L2’s brickwall peak limiter acts as a ‘protected gain’ control. It lets you
increase the gain as you choose, as well as define a maximum output ceil
-
ing which the signal is never allowed to exceed. The L2 limiter only reduces
the applied gain locally when it anticipates that a clip is about to occur.
By transparently controlling signal peaks, the entire level of the source can
be raised several more dB, resulting in a higher average signal level. For
real time processing, the normalization process of disk-based systems is
not possible, as it is nearly impossible to know the highest peak of a source
in advance. The L2 prevents overshoots by utilizing a “look-ahead” tech-
nique, which lets it anticipate and reshape peaks in a way that produces a
minimum of audible artifacts (at the expense of introducing a 64 sample
throughput latency).
Details on using IDR
TM
(Increased Digital Resolution)
In the MaxxBCL, all processing is done at high resolution; in this case, with
48-bit precision. When processing at a high bit depth, you must decide
how you want to finally reduce this long wordlength for the output. Any
digital signal processing that alters the original digital data (mixing, gain
changes, EQ, dynamic processing, etc.) generally increases the number
of bits required to represent the signal. Conventional truncation results in a
loss of signal-resolution each time the signal is processed.
The solution is to properly dither and noise-shape a signal each time the
wordlength is increased and then reduced (such as nearly every digital
signal process will require). This is to maintain the best possible quality at
the reduction phase.
Inside the MaxxBCL, all data is required to use 48 bits to represent the sig
-
nal, even when MaxxBCL is processing a 16-bit input signal. Therefore, even
when the output from the MaxxBCL is 24-bit, the IDR™ system is used and
captures up to 27 bits of perceived resolution for the output signal. Truncat-
ing the 48-bit internal signal at the final output will result in a rounding error
which produces distortion at low signal levels and a loss of digital resolution
that can never be recovered. Needless to say, IDR™ is much more critical
in the case of 16-bit output into 16-bit digital devices, but it is also benefi-
cial when the output is 24-bit.
maxxBclManualFinal.indd 19 4/19/05 6:25:02 AM