On 10/28/2018 1:35 PM, Les Cargill wrote:
A digital mixer cannot have actual zero latency. It can have
sub-millisecond latency. Well below anything like the Haas limit. I
used to use the Fostex VF16 as a gig mixer and nobody noticed anything.
For sound through the air, small amounts of latency are no problem at
all. Think about playing an electric guitar and standing 5 feet from the
amplifier. Sound travels about 1 foot in a millisecond, so, assuming
there's essentially no latency between when you pick a string and when
the sound gets to the loudspeaker - a reasonable assumption given the
speed of electricity - that sound won't reach your ear until 5
milliseconds after you've picked the string.
Where latency is a problem is when there are two paths for a sound to
get to your ear. When you're speaking or singing, the sound of your
voice gets to your ear through two relatively short paths, pretty close
to equal length - one from inside your throat directly up to your
eardrum and the other through the air from your mouth to your ear. But
when you put headphones on, the situation changes. You still have the
"internal" path, but the external path is replaced by what feeds the
headphones. When that's on the order of 1.5 to 3.5 millisonds and pretty
close to the same SPL at your eardrum as the sound through your throat,
you've created a comb filter that puts several notches right in the
speech range and your voice sounds un-natural.
The engineer in the control room or the rest of the band will hear your
voice just fine, it's only the singer who's affected, and it's only when
he's singing. When he hears the playback (if he can perform well,
hearing this odd version of his voice) it'll sound fine to him.
Many people tell me "I've never hear that" and I believe the reason is
that they have the headphone volume enough higher than the internal
volume so that the notches aren't deep enough to create much havoc. If
the singer starts out loud and then asks to be turned up, he'll never
hear it. But it's particularly annoying to spoken word artists who only
want enough volume in the headphones to know that things are working -
which is why direct analog monitoring is usually the setup when there's
someone in the vocal booth.
You can simulate this easily in a DAW. Record a voice track, then copy
it to another track and shift one track by 1 or 2 milliseconds. Set them
to equal volume and listen to one, then both summed.
"Zero latency" for things like the Focusrite Scarlett series just means
"minimum possible latency" - it has a very minimal digital mixer built
in.
I call that "Zero latency for large values of zero."
--
For a good time, call
http://mikeriversaudio.wordpress.com