> Peter L. Peres" wrote:
>=20
> On Tue, 19 Nov 2002, Bruce Douglas wrote:
>=20
> *>The best links I know of on the web are:
> *>
> *>http://www.philrees.co.uk/articles/timecode.htm
> *>http://www.inforamp.net/~poynton/notes/video/Timecode/index.html
> *>
> *>I made an LTC reader using a 628.  The approach I used was to rot=
ate
> *>the input bits into ten registers until the 16 last bits agreed w=
ith
> *>the SMPTE identification bits and then decode the other registers=
 to
> *>HH:MM:SS:FF.
> *>Since both up and down transitions are significant I used the int=
errupt
> *>on state to read the signal.  I don't know if this is the best wa=
y but
> *>it worked for me.
> *>
> *>I was toying with the idea of making a LTC generator but was put =
off by
> *>the fact that this would require a precision of at least 1 frame =
(1/30
> *>second) in 8 hours (1 part in 864000) and I think this is beyond =
the
> *>precision of even temperature controlled oscillators.  If I'm wro=
ng
> *>about this I'd be grateful if you could let me know.
>=20
> You are not wrong about this but if the LTC signal does not live by
> itself but is connected to some video source then you can resync on=
 the
> video source. The LTC is not meant to be a real time clock that exc=
ceeds
> the precision of the video source, the latter being +/-10ppm in the=
 best
> of cases. There is also the problem of interleaving LTC with frame
> boundaries correctly, and that requires that the LTC follow the vid=
eo
> frames or vice versa.
>=20
> Also if anyone told you that LTC from studio equipment that is not =
locked
> to a rubidium clock is accurate to 1.6ppm, then imho they lied ;-)

http://www.nagrausa.com/NAGRA_IV.htm
"Time code generator stability:

+/- 1 ppm from -10 to +40=B0 C corresponding to a variation=20
^^^^^^^^^^
of +/- 1 frame in 9 hours of operation at 30fps or in 11 hours
at 25fps (+/- 2 ppm from -10 to +60=B0 C) "

Sure looks to me like Peter is mistaken (unless the unit was too hot
to touch. Proteins denature at 42C). The old workhorse Nagra_IV
had a 1ppm spec. Any wonder then, that they were so popular?


http://www.locationsound.com/98summer/newprod.html
Nagra ARES-C with Time Code=20
 The Master clock accuracy is +/- 3PPM.=20

I guess the specs got looser when everyone started using
frame synchronizer, but didn't both to delay the audio to match.

Everything you could possibly want to know about film motors and
accuracy.
http://members.aol.com/npr16mm/motors.htm
Cable sync was how the first "portable" double-system sync sound syst=
ems
operated. You had to have a cable running
between the cameraperson and the sound recordist at all times.


Robert

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics