[time-nuts] GPS for Nixie Clock
jimlux at earthlink.net
Sat Jul 16 09:32:30 EDT 2016
On 7/15/16 6:23 PM, Chris Albertson wrote:
> I did write that it's useless to have a visual display that is three
> orders of magnitude better than the human perceptional system and was
> corrected that such a display could be used for film based
> photography. That is true. But that just adds even more problems
> like making certain the display never changes while the camera shutter
> is open. These old camera time loggers were hooked up to the shutter
> release. I think they captured the time of day when the shutter opens
> and hold it for the duration of the exposure. I have some old ones
> that I can check, but I'm certain they did not change while the
> shutter was open. They did not light up at all when the shutter was
> closed. If they did change without regard to the camera shutter then
> on order 1% of the shots would capture the increment and with a
> 7-segment number it would be unreadable. But this never happened.
WHen I was working in the film/tv business, I made a lot of devices (or
more properly wound up *modifying* existing devices) to take a camera
sync pulse so that we could be sure that the display was static while
the shutter is open, or that the xenon strobe or blinky lights would
fire at the right time.
Typically, you need an adjustable delay box with two settings that you
set empirically on set. You get a pulse for the shutter sync (it's 15
years ago, and I don't remember the details), but you need to have two
possible delays: the camera operator's viewfinder is open when the
shutter is not = that is, the optical path is essentially switched
between the view finder and the film. 555s were my friend.
This is used to great advantage on closeup or in remote control shots
where they use a laser pointer aligned the the optical axis, and only
turn it on when the viewfinder is open. You can "see" if you're
pointing the right direction, but the laser is off when the shutter is
open to the film. I imagine by now there are fancier versions that
project frame lines or corners and such.
You also sometimes need to adjust the current or multiplexing rate to a
LED display so that the brightness doesn't change with changing shutter
duration or phasing. You definitely need to sync the multiplexing with
the shutter or the display will either be partial, or will have a
strange cyclical brightness variation.
There's a whole industry of producing 24 fps TV and computer monitors so
the display "looks" ok when filmed at some rate. I used to have a bunch
of "genlockable" VGA display cards that could be put into a PC and
synced to some supplied sync signal. I wrote a fair amount of little
utility programs that would poke the registers on a video card to get
specific vertical frame rates and you'd hope the production staff had
bought monitors that could sync at 24 or 48 fps, or some other oddball
rate because they were shooting slo-mo at 120 fps or something.
I think today, with much improved automatic compositing and offline
editing, they just paint the screens green or blue on set and composite
the video information in later. Even with a camera move in the shot, the
compositing operator would mark the corners of the screen in a few
frames, and the software would do the rest.
More information about the time-nuts