Lang Elliott wrote:
> Thanks Walt. I was afraid this was the case.
It might be possible to modify the recorders to bring the clock signal
out from one to run into the other. Or to use a external sync clock for
both. But that will take someone very familiar with them.
Walt
________________________________________________________________________
________________________________________________________________________
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>From Tue Mar 8 18:22:56 2005
Message: 21
Date: Wed, 06 Nov 2002 23:17:22 -0500
From: Walter Knapp <>
Subject: Re: Lang asks Tech Heads
Marty Michener wrote:
> At 08:55 AM 11/6/02 -0500, you wrote:
>
>>I have a question for all the tech heads:
>
> Hello Lang!
>
> I'll let the head tech (Walt) answer most of this, but my take is: the
> synchronization you will probably be looking for will not be there. Ears=
> can hear, according to experiments in the 1960's, time differences of abo=
ut
> 10 to 50 microseconds, depending on to whom you talk. The sharpest coin=
> click you could record would not give sufficient coincidence to make the=
> later synchronzation possible, especially as it is recorded in digital
> form, notorious for ruining arrival times.
You really need a marker pattern that's on the order of just a sample or
two. The better way to do it is a small pulse generator wired into all
the mic inputs. That avoids all the phasing problems of recording a
sound. For 44khz sample rate, the sound is moving about 1/3" during each
sample period. That's pretty precise locating for recording. The one
external way would be to tap the housing lightly with some object that
makes a real sharp sound, that is if the housing is continuous.
But if you go off and record possible clickers, you will find that their
sound is pretty drawn out vs sample length.
How did I get this new title? Fumbler tech might be more like it. I'm
just a biologist.
> The head tech has already made several posts about drift, regarding his
> head Mac and his PortaDisc and/ or DAT recorder.
> They might drift apart as much as a sample every ten to thirty seconds, I=
> would guess.
Difference between my Portadisc and my G4 is very stable at one sample
every 2.76 seconds. The G4 is the one with the slower clock, and the
amount is about the same as the amount it's internal clock drifts
against time standards. Which would imply that the Portadisc is running
at a rate close to the standards. I just reset the time in the Portadisc
a few days ago to match and it had drifted about 3 minutes since last
spring. The difference between those two is enough to cause a beat
frequency that you hear as a varying sound level. Running in a surround
system, it would be enough to hear that way too.
I've kind of assumed that recorders probably use crystals that are more
precise than computers. But all are spec'd with a plus/minus in the specs.
Walt
________________________________________________________________________
________________________________________________________________________
|