3/23/2011 10:46 AM
Focusrite’s Pro 24 DSP is just one of many audio interfaces that includes a mixer application for zero-latency monitoring.
I see a lot of audio interfaces that advertise
“zero-latency monitoring.” I thought there was
always at least some amount of latency when
using DAWs, so how do they pull this off? And
why would anyone use other interfaces if some
offer no latency when monitoring? I get different
answers from different people, so I thought I’d
indeed “ask EQ.”
Troy “TJ” Jeffers
Las Vegas, NV
EQ: You’re correct, there’s always some degree of latency
when using a DAW and monitoring through a computer,
but understand this isn’t necessarily a constant amount,
nor does it relate solely to the interface—a faster computer,
simpler projects, and better drivers are just some factors
that contribute to lowering overall latency.
Latency occurs because audio is a realtime process,
but computers are constantly being distracted by other
tasks—monitoring keystrokes, checking ports, etc. So,
storing some audio in a memory buffer means that if the
computer can’t deal with audio at any given moment,
there’s some in reserve that can be streamed to keep the
audio flow going. The amount of audio you put in reserve
translates to the overall amount of latency.
Interfaces with zero-latency monitoring get around
this problem by bypassing the computer altogether,
typically through a bundled mixer application that allows
patching the interface inputs directly to the outputs.
However, the biggest disadvantage to zero-latency
monitoring becomes clear if you need to hear the sonic
effects that plug-ins impart to a sound, such as when
recording guitar with amp sims. With zero-latency monitoring,
you’ll hear the guitar’s dry sound that feeds the
input, not the sound processed through the amp sim.
(Incidentally, zero-latency monitoring is seldom true
zero-latency monitoring, as putting a signal through A/D
and D/A conversion typically adds about 1.2ms of
latency at 44.1kHz. But most would consider this negligible,
because 1ms is roughly equivalent to moving a
foot further away from your speakers.)
Some interfaces incorporate hardware DSP to provide
effects that don’t require plug-ins. For example, if
the interface includes reverb, then it’s usually possible
to add reverb to the monitored signal so a vocalist can
hear their voice with reverb—without having to use a
reverb plug-in, or recording the voice with reverb.
Bottom line: Use zero-latency monitoring if you don’t
need to hear the results of any processing applied by
the computer; otherwise, do what you can to minimize
latency, with the main improvement coming from (sorry
about this) upgrading to a more powerful computer.
Ask EQ a technical audio-related question, and EQ will
answer it. Send it to EQeditor@musicplayer.com.