[time-nuts] Measuring TV delays

Magnus Danielson magnus at rubidium.dyndns.org
Fri Jan 3 13:26:50 EST 2014


David,

On 03/01/14 16:28, David J Taylor wrote:
> From: Magnus Danielson
>
> There are several potential time-signals, but I would say it would be a
> bit hard to verify their traceability, which makes the exercise difficult.
>
> The MPEG-2 Transport Stream have PID timestamps for every channel.
> However, there is really no good TAI/UTC traceability in them. Also,
> re-stamping is used to cover up for delay variations.
>
> VITC might be present, and might provide local time, but for NTSC
> countries there is a frequency error only partly covered in the
> drop-frame compensation which causes a deviating display-time from
> actual local time.
>
> Caption can also carry time, but there is no guarantee it has a sane
> source.
>
> Cheers,
> Magnus
> ===============================
>
> Many thanks, Magnus.  There seems to be a lot of scope for
> experimentation and discovery here!  Excellent point about traceability!
>
> David

Thanks David, if you want to make delay measurement, you need to make 
assumption about the correctness of time and I'm trying to point out 
that there is a limit to how accurate it can be, and hence the 
measurement can be fairly bogus. Essentially, in the TV business they 
don't care very much about accurate time that much. If you have a hint 
from your watch when to expect a particular program, then it's OKish.

If they did care, it would be not too hard to fix it, but it is becoming 
harder than it used to be.

Cheers,
Magnus


More information about the time-nuts mailing list