Advertisement
Application Guide
7
to as propagation delay or "latency", and is typically only a few thousandths of a second.
However, this slight delay becomes significant even at very small values. There are two
potential problems with latency. The first is where the latency becomes so great that there is a
perceptible echo between the original sound, and the amplified sound. This manifests itself as
a problem mostly for the performers, where it becomes increasingly difficult for musicians to
keep in time and tune with each other, when relying on a monitor mix which has an excessive
delay. There are no hard rules for this, different musicians have varying degrees of tolerance
to latency, the most critical scenario is where in-ear monitoring is used. Overall latency
seldom causes problems for the audience, as the speed-of-sound propagation delay of the
audio through the air of the venue will be many times greater than any digital processing
latency. The second, and more complex issue with latency, is that of differential latency.
Differential latency is where the same audio stream is routed via multiple paths which have
differing delays. Audio engineers commonly route audio by different paths, apply processing
to those paths, then re-combine the audio at a mix bus. Attempting this with any digital
processing in the signal path will cause comb filtering due to audio samples arriving at
different times. Comb filtering means that at certain frequencies, by a process of phase
cancellation, some audio information is irretrievably lost. The audio spectrum of the missing
information is dependant upon the time differential, and once the data is lost, it is impossible
to correct for this. The difference may only be fractions of a millisecond, but that is enough to
create undesirable audio artifacts which will have a profoundly detrimental effect on the audio
quality. This loss of audio quality is common in many digital mixers which do not have any
form of delay management.
In a recording environment, this latency can be managed by buffering the audio, which means
measuring the longest delay imposed upon an audio path, and applying the same time to all
other signal paths. This ensures a phase-coherent mix. Through-system latency can
accumulate up to many milliseconds, which, in a recording environment need not cause any
problems, as long as the musicians are monitoring pre-processing and in real-time. This
becomes more difficult in live sound applications, as everyone, musicians, mix engineers and
audience are all monitoring post-processing and in real time!
As a stand-alone mixer the VeniceF has the same "instantaneous" signal path as any other
pure analogue mixer. When using a computer connected via FireWire, configured to record
via the VeniceF's direct outputs, groups, auxes, matrix and masters, the same latency is
present in all paths, so the result is a phase-coherent multi-track recording.
When using the computer to provide FX using the VeniceF's aux sends and its stereo inputs
as the FX returns, because the FX are all time-based (reverb, delay, chorus, flanging etc.) the
only consequence of any additional latency will be that the pre delay on any reverbs, and any
echo effects, will be delayed by the latency of the FireWire driver. This additional delay will not
create any issues during normal operation. However, if the computers processing is used as
channel inserts, the operator must be aware of the possible consequences of recombining
any of the "dry" processed audio from the FX with the original analogue signal. This could
cause comb filtering, and subsequent audio quality issues, similar to those experienced when
using digital mixers or plug-ins which do not feature automatic delay management systems,
as used in the Midas PRO series and XL8 digital mixers.
To avoid these undesirable differential latency effects, it is advisable to apply identical signal
paths to all audio channels. When using computer-generated processing as inserts on the
VeniceF
Application Guide
Advertisement
