SCOUG-HELP Mailing List Archives

<< Previous Message << >> Next Message >>

Date:	Tue, 30 Sep 2003 10:07:55 PDT7
From:	Peter Skye <pskye@peterskye.com >
Reply-To:	scoug-help@scoug.com
To:	scoug-help@scoug.com
Subject:	SCOUG-Help: time accuracy (was: Mail server down??)

Content Type: text/plain

=====================================================
If you are responding to someone asking for help who
may not be a member of this list, be sure to use the
REPLY TO ALL feature of your email program.
=====================================================

Steve Carter wrote:
>
> I'll wager lunch that the "head honcho" made the
> change to quiet down all the duck quacking in his ear!

'Tis true. All that quacking is now properly organized by timestamp.

> >Your old timestamp is still in the header
> > -- take a look and you'll see it.
>
> It's kinda neat to see how far off SOME people's
> clocks are! Mine gains 2.5 seconds/day!

Computer clock chips aren't as accurate as some might think. I have one
machine that gains almost two minutes per day.

You should be running a Network Time Protocol (NTP) client on your
machine to keep the clock accurate. (Steve, I thought you were running
TIME868.) My favorite is OS2NTPD because it keeps my clock more
accurate than any of the others. For some reason it's never been
uploaded to Hobbes but you can get it from the European equivalent LEO:

http://ftp.leo.org/pub/comp/os/os2/leo/tcpip/systools/os2ntp12.zip

I believe eCS comes with one. Personally I like the
constantly-synchronized approach of OS2NTPD rather than the
occasionally-synchronized approach of the others. Don't forget to use a
separate non-PM NTP program (I use DAYTIME) to initially set the clock
on bootup from your startup.cmd file (*not* your Startup folder) or from
your CONFIG.SYS file.

Now then, to start a separate thread if anyone is interested, the clock
chips on sound cards aren't accurate either. This leads to a couple of
interesting problems. First, if you record a musical "A" note (440 hz)
with one card and play the file back through a different card then the
frequency won't be the same (sort of like a slightly off-speed
turntable). Second, if you're listening to a radio broadcast over the
internet then you will eventually either underflow or overflow the
server's buffer size and lose sync with the broadcast. I've read a
number of threads about people being disconnected after listening to
internet radio for a while; what they don't realize is that they've
simply lost sync with the server because their sound card is running
slightly off-speed.

At the commercial (radio & tv) transmitters we used
temperature-controlled (an oven within an oven) crystals to stay on
frequency, with monthly certified frequency checks against the
government's WWV frequency broadcasts. A much simpler calibration
technique is to record the 60 Hz frequency from your wall outlet (it's
very accurate because it's synchronized with the power grid) and then
check the file and see if each cycle takes 44,100/60 = 735 samples. You
can get recordable 60 Hz from a low-voltage "wall wart" power module.

- Peter

=====================================================

To unsubscribe from this list, send an email message
to "steward@scoug.com". In the body of the message,
put the command "unsubscribe scoug-help".

For problems, contact the list owner at
"rollin@scoug.com".

=====================================================

Return to [ 30 | September | 2003 ]

SCOUG, Warp Expo West, and Warpfest are trademarks of the Southern California OS/2 User Group. OS/2, Workplace Shell, and IBM are registered trademarks of International Business Machines Corporation. All other trademarks remain the property of their respective owners.