|
|
Posted by Rick Merrill on 10/08/07 16:13
Richard Crowley wrote:
> "Rick Merrill" wrote ...
>> Richard Crowley wrote:
>>> "Rick Merrill" wrote ...
>>>> My concern is that the generator is not keeping very
>>>> good timing so the audio begins to lag over the course
>>>> of a couple of hours recording. AFIK there is no time
>>>> sync on the audio, just on the video.
>>>
>>> No equipment, even professional will hold sync over "a
>>> couple of hours". Or even over one hour. They is why
>>> we use gen-lock and timecode, etc.
>>
>> Call me suspicious, but if "gen-lock" signal is off by
>> 0.1% then after 10 hours of recording there will be a 1%
>> cumulative error.
>
> First, "gen-lock" is not a signal. It is a method of interconnecting
> equipment so that the sync signal of the "master" unit is
> running the "slave" equipment. The slave equipment is said
> to be "gen-locked" to the master.
>
> And you are correct. If the sync signal is 0.1% off, then it
> may be off by 1% after 10 hours. But remember that most
> equipment drifts in BOTH directions so it is PLUS OR MINUS
> 0.1% and unlikely to have accumulated all in one direction.
>
> But all equipment drifts to a greater or lesser extent. That
> is why if you need to keep separate video and audio
> recorders in sync, you "gen-lock" them together so that they run at the
> same rate.
>
>> It is the accuracy of the gen-lock that comes from the black-burst
>> generator and that's what I wanted to know.
>
> Most equipment is likely accurate to better than 0.1%
> But likely NO equipment is absolutely accurate (except
> occasionally by accident). The reason is that it is just
> not worth the cost to make something that is absolutely
> accurate. Nobody relies on the absolute accuracy of the
> sync signal to lock together two different pieces of gear.
>
> In places where agreement with wall-clock real-time IS important (such
> as broadcast on-air operations, master-
> control, etc.), they tend to use (nowadays) sync generators
> that are themselves synced to a secondary reference such
> as the constellation of GPS satellites, or WWV, etc. (in
> the US, likely similar references in other places, etc.)
>
>> In other words, does anyone ever check the accuracy of a genlock
>> signal (e.g. one that was purchased on the cheap from ebay)?
>
> Sure, most equipment of that kind can be "calibrated" by a
> technician with the appropriate experience, information, and
> equipment. Of course, it will almost certainly cost more than
> you paid for it on eBay, which is why it is rarely done.
>
> If you are looking for a solution to keep separate video and
> audio recordings in sync over two hours, the absolute accuracy of the
> sync generator is NOT the solution. Even if you had an unlimited budget
> and could buy the most expensive equpment
> ever made, people don't run open-loop and depend on the
> absolute accuracy of the equipment. They interconnect the
> equipment together ("gen-lock") to ensure agreement of speed.
>
> Or else they just live with whatever "slippage" there is and then
> simply correct it in post-production on their NLE timeline. That
> is far and away the cheapest and simplest solution.
>
Thanks very much for the tute - that gives me a couple of avenues
to investigate, namely to find out how the audio recording uses
the genlock we feed it.
On the output side I was surprised to find that the MTS-4 Compander also
takes a genlock signal from the video frames - any comments about that?
Navigation:
[Reply to this message]
|