building a multicamera remote trigger with shutter delay timer

[metter]

Hi

I am completely oblivious to this, so pardon if this does not sound very technical right now.
I was wondering if anybody could tell me the scope of this little project I am working on.

For an arts project, I need a triggering device that can trigger 150 DSLR cameras at the same time and then closes the shutter on one camera after the other with a preset delay interval. The cameras are triggered with a shutter release cable (basically an audio cable with a sub mini audio jack). The whole thing would also need an additional timer that delays the above shutter sequence after the trigger button is pushed.

So basically I assume this would be a box with 150 sub mini audio connectors at the back. And on top would be a trigger button, a dial for the delay until the closing sequence starts and a second dial for setting the time interval between each camera.

Cameras would be set to bulb mode, allowing to control "shutter open" and "shutter close" by closing the circuit of the shutter release cable.

Now I obviously cant build such a thing, but I would be very helpful if somebody could tell me what to expect in material costs and labour that is required to build such a thing.

any help?

thanks a lot!

 

[arthur0]

Hi.

I’ve seen that there are commercial PC-based solutions costing many hundreds of $, but since you asked in a DIY forum, I’m guessing you don’t want to spend that kind of money and will opt for a custom (hopefully less expensive) solution.

I thought your project was interesting and waited until someone, more experienced than me in electronic design, would post a solution. Since this hasn’t happened yet, I tried to look into it myself, although, again, this is not my primary area of expertise.

If I understood correctly your description, you want to open the shutter of 150 DSLRs, expose for a preset interval (T1) and then begin to close their shutters one by one, in sequence, after a preset delay (T2). (I must admit that having increasing exposure times doesn’t make much sense to me. I’ve only seen “bullet time” effect.)

If I was to do this, I’d do it around a uC (MSP430 or an 8-bit Freescale, for instance) with a 16-bit (or more) timer with a corresponding PWM output, gated by a IRQ trigger-button input. The timer will be used to generate 2 PWMs, first with period T1, then T2 (corresponding to the initial exposure time and delay, respectively).
This waveform will be fed as clock to all of the needed 19 serial to parallel shift registers (a good candidate being 74F322). Their output will (either directly or through inverter buffers if the logic requires), control the triggering input of the cameras.

All the timing settings can easily be fed (via rotary encoders, maybe) to the uC, which will also take care about resetting the register to initial conditions.
Neither the hardware nor the software for this project is too complex. It’s its sheer size that’s scary.

I hope you’ve got an idea about how the bill of materials might look and the amount of work involved so you can do some cost estimates yourself.

Arthur

 

[metter]

Hey Arthur

Yes I started to see tumbleweed around here too. Thanks a lot for your reply.
I have been milling on this project for ages now. But it is becoming more and more plausible now.
I was looking for the above mentioned PC based solutions, but couldn't find any. What are you referring to? The only thing I found was a guy turning his shutter cable in a serial port cable that could be triggered via PC. Here. But that only solves my problem by 1/150th. But it was good to see that the triggering mechanism actually is very simple, raising hope that the above mentioned trigger box would not be too complicated to build. I was even looking at a strictly mechanical solution! With rotary timers etc.
Yes you got the basic functionality right. I would say though that T2 is not so much of a timer, but an interval setting. Like this:
if
Sx = camera shutter time
T1 = shutter close delay
T2 = closing interval
then

Sx = T1 + (x*T2-T2)

Thanks for the explained approach. Even though I understood almost none of it (I have zero knowledge of electronics), it did let me run some quick google search for the components. That was very helpful as I already have a list of most of the other cables, adapters etc needed for this. And hey - this actually looks reasonable!

The only thing I am still oblivious about is what it takes to assemble something like this. I know it should not be underestimated as there are 150 connectors involved!
But still - for somebody who actually knows what he is doing - and given a box with components. A days work? A week?

I need to cost this up because it obviously needs to be bankrolled first.

 

[arthur0]

Hi,

It seems I got your requirements right. (Although, I must say, I still don’t understand the system’s purpose. If you make a movie from the sequence of pictures, won’t it look like the picture is fading-out (assuming the aperture setting is the same on all cameras)?)

When I say “timer”, I’m referring to the uC’s hardware timer that needs to be configured by software to count a specific time interval.
You can understand my approach by imagining that (starting with all the shutters open) the closing of the shutters is a wave spreading from the first camera to the next. That’s what the serial to parallel shift registers (connected serially to each other) will be doing. The rate of the advancing front of this wave is determined by your T2 interval and will be provided as clock signal by the uC to all the registers. The shorter the interval T2, the faster the clock signal, of course. T1 is obtained by configuring the uC to hold the clock signal for the preset amount of time (counted by the same or another internal timer).

The commercial PC-based application I’ve seen is here. To the cost of that program you have to add the cost of the hardware: 5 computers and about 19 USB-to-digital-output boards. This amounts to well above $20000! (Ridiculously expensive if you asked me. I almost want to commercialize my own, hihi.)

The solution I’m proposing requires the manufacturing of one (or more) PCB, but once you got that and all the needed parts, I’d estimate it can be assembled in a couple of man-days. Together with the uC firmware it should be doable in a man-week if the motivation is high .
At the same time, this design can be laid out with some degree of flexibility so you could achieve a lot of other interesting multi-camera setups.

Arthur