Re: Synchronizing camcorder with telemetry

From: Jim Stewart (jstewart_at_jkmicro.com)
Date: 07/08/04 

Date: Thu, 08 Jul 2004 04:03:33 GMT

Lewin A.R.W. Edwards wrote:

> Hi all,
>
> I'm trying to think of a way of putting a timestamp onto an analog
> camcorder recording so that it can be synchronized with a vehicle log,
> and would appreciate comments and suggestions. The end goal is to
> recover a video stream that can be played back with realtime telemetry
> alongside it (after post-processing).
>
> Background: This camera is one possible payload in my submarine. The
> payload bay has a 9600bps "kinda-RS232" interface to the vehicle's
> main computer, and the idea is to make the protocol on that interface
> as general as possible to permit different things to be installed
> there. I've developed a "generalizable" imaging device controller on
> an AVR. It talks RS232 and has eight open-collector outputs designed
> to drive the inputs of the rotary mode switch often found on
> camcorders, and also the record start button. It also has two inputs
> designed to detect the state of LEDs or other indicators on the camera
> (eg. to detect end-of-tape by determining that the camcorder has
> switched off the REC LED), a CdS sensor and a driver that can switch
> the unregulated system battery voltage into a lamp. So in theory, all
> I have to do when adapting this ckt to other camcorders is to
> reverse-engineer the switch pinouts and maybe change the firmware in
> the tiny26L.
>
> I still have a couple of GPIOs and LOTS of unused processing time.
> What I'm thinking is that I can cut off the electret mic on the
> camcorder, put an appropriate analog buffer on the audio input, and
> use the aforementioned imaging controller to PWM timecodes (or even
> actual metered data) onto the audio track. Has anyone done something
> similar to this? Is the audio speed control reliable enough for this
> to work on an average camcorder?
>
> Another idea was to flash an LED to reflect off the window in front of
> the camera, as a log-sync marker. The timestamp of the LED flash event
> would be recorded in the vehicle log for later correlation with the
> recording. This relies on the camera being able to maintain reasonably
> constant frame timing. The advantage of this system is that I can
> still get the actual audio data from the environment around me. I
> anticipate that the total recording time will not exceed two hours,
> and up to 2.5 seconds' skew between telemetry and video is acceptable,
> though less is better, naturally. Is this achievable with, say a VHS-C
> camcorder in SLP mode?

You might want to consider writing SMPTE timecode to
an audio channel. The technology is ancient, robust,
and well documented. In particular, it is self-clocking
in such a way that it's readable over a wide range of
playback speeds.

If you implement the standard frames/seconds/minutes/
hours/ fields, you have 32 user bits left. And that's
at 30 times a second. If SMPTE compatability is not
an issue you could just use the modulation scheme and
get 68+ bits per field.

http://www.philrees.co.uk/articles/timecode.htm

Quantcast