[Vicktor] has always been fascinated by photographs of lightning and decided to try his hand at capturing a few strikes on his camera. Every time he attempted it however, he didn’t have much success. Instead of trying to operate his camera manually to take the images, he decided to build a lightning trigger that would do it for him.
His circuit uses a large photodiode to sense when lightning strikes, triggering the camera via a hacked shutter release cable. A PIC micro controller is used to adjust the sensitivity of the device, as well as to send the actual trigger signal to the camera. His circuit is connected to the camera via a pair of opto couplers to ensure that his circuit cannot cause any harm to the camera.
When the box is powered on, it enters a calibration mode where the user can adjust the circuit to compensate for whatever amount of ambient light is present. Once armed, the box waits for a sudden change in ambient lighting, sending the exposure release signal to the camera.
A schematic is available on his site, and he will send you the code he use on request. There is currently no video of the trigger in action, but hopefully we’ll see one soon.
If you’re interested in seeing some other remote camera triggers, check out this one made from air freshener parts, and this one which uses lasers.
I also heard it’s possible to make a trigger using a radio receiver, I which case it’s possible to detect a lightning like a second before it happens!
Aww I was hoping for some pics I too enjoy lightning.
You don’t have to trigger the camera when there is a flash of lightning to take a good picture. Just use a long exposure of 1-2 seconds. It takes a bit of experimentation to get the exposure correct. I’m sure a trigger works better but for anyone who wants to try lightning photography this is an easy and accessible method. Look at one of my results : http://picasaweb.google.com/the.antonio.lucero/Thunderstorm102410#5532556991287719410
Hadn’t thought about opto-couplers for my own version, I should probably add that… My version will be quite a bit smaller (simpler, no adjustment knob) and also acts as an IR-remote. Now, if the damn postal system could do its job so I could finish the thing (been waiting a couple of months for two separate packages now *grr*).
Radio receivers detect lightning you can’t see or hear. Detecting lightning a second ahead of a nearby stroke sounds like too long. Although there is a sequence of discharges, it’s shorter than that. You might be able to detect some of the less powerful discharges before the main one, But how would you distinguish them from distant bolts? I don’t know if there’s any time or frequency difference.
Even an optical detector is going to need to be fast to photograph a bolt while it’s still in progress.
You can do some science here.
It is also possible to use CHDK and some motion detection script for lighting photography. The lag is about 200ms.
If you have a Canon point and shoot, you can start lightning photography without extra HW using the CHDK firmware, discussed here before: http://hackaday.com/?s=chdk
The best lightning photos I’ve seen use anywhere between 10-30 second exposure, so wouldn’t need a device like this.
Still quite cool though.
i would not have used that a-d chip it is very laggy i found out using one as an encoder
but i got rather nice lightning shots with just hitting the shutter http://biozz.deviantart.com/gallery/?q=lightning#/d2og21c http://biozz.deviantart.com/gallery/?q=lightning#/d2og5li http://biozz.deviantart.com/gallery/?q=lightning#/d1n8gjz
Alternatively, I have heard of people just setting a manual exposure. That would only work if you are in a secluded area though…
You could do away with the calibration step by using a simple highpass filter (analog or digital).
To get lightning images, you simply need to set the camera on a tripod, (or some other solid surface) set the aperture to f16, then leave the shutter open. (while pointing the camera in the general direction of the lightning.)
I’ve done it…but the area I live isn’t conducive to great weather photos because of the flat terrain and buildings. But it does work…provided it’s not daytime. :)
Yeah, I wouldn’t shoot lightning like this. I use a tripod and a remote shutter release (to avoid jostling the camera while the shutter is open). Set the shutter speed to “bulb”. Compose. Set aperture by trial and error (depends on how long between strikes / ambient light levels / how impressive you want the strikes to be). Open the shutter. Wait for a lightning flash (or as many as you want to be in the picture). Close the shutter. Repeat until you get rained out or the lightning starts to get too close for comfort.
The problem isn’t going to be lag in your detector, but lag in your camera. Lightning only lasts a few hundred milliseconds at most and anything less than a good SLR will have a longer shutter lag than that.
Which is why locking the shutter open and pointing it in the general direction works best. I’ve done it tons of times that way. You also have the bonus of seeing the rest of the surrounding terrain instead of just a flash of lightning on a black field.
If you pre-focus and shoot in manual shutterlag typically ranges from 84ms for a D3000, to 41ms for a D3. This would imply that you still get some amount of lightning in the exposure, but not as much as if you’d shot properly to begin with.
One thing that really gets on my nerves is designs like this. Opto coupler with a comparator and a voltage reference would have done the job nicely. Instead now we have a micrcontroller, the microcontroller introducing it’s own sample time, and program delays.
Aside from the fact that the problem shouldn’t exist, the solution is the slowest and most expensive possible design. A very neat looking design, but slow and expensive none the less.
I’m the original designer of this circuit (http://fototipp.hu/light2). First of all, the original goal of this circuit was to make possible lightning photography in daylight. This is why it has the OPT101. Yes you can do it with a number of op-amps and some timers, but it’s easier and more convenient to do with a microcontroller. The delay added by the controller can be measured in a couple hundred nanoseconds and this is irrelevant.
When detecting lightning for photography, you have to detect the very first event which called the
“stepped leader”. This event, which is usually takes about 100ms, precedes the lightning strike, and emits light mostly in near infrared. The circuit was designed to detect this kind of events.
Yes, indeed, you will need a good DSLR to get a lightning, however with a correct detection of the stepped leader, you will be able to capture most of the lightning, since an average DSLR shutter lag is less than 100ms.
I doubt that these pictures could be taken with any other technique:
http://fototipp.hu/galleria/sefi/villamok/villam-12
http://fototipp.hu/galleria/sefi/villamok/villam-13
http://fototipp.hu/galleria/sefi/villamok/villam-9
@Sefi – as an avid amateur photographer here, with those photos, WIN! I am so building one to use with my DSLR and possibly my 35mm.. got a few rolls still good of some b+w film i’ve been looking for a reason to use.
I can’t ever imagine the need for something so complex. Just a photocell or magnetic pulse detector aimed in the same direction as the camera driving an op-amp threshold detector driving an optocoupler would be far more than enough. I doubt it would cost a few dollars in parts plus an hour or two of time.
I agree with the other person who said the real problem is shutter response time of the camera.
What really needs to be done is have a camera that takes one photo after another but can be programmed to save the photo taken during the trigger pulse.
My home security system does that. It constantly processes video frames and records any frame where motion or rapid image change is occuring.
Old post, but I built a similar system several years ago, using a fast PIC running at 40MHz. The nice thing about using a micro is that you have a lot more control over calibrating out the ambient light, which makes day time captures that much easier. You could use a high pass filter as someone suggested, but you can implement and calibrate one in software a lot more easily and quicker. I also have a delay setting, which locks out the camera for a selectable amount of time, because you will get a lot of false positives from strikes that are just out of frame. You miss some real strikes, but the false positives happen way more often and can fill a memory card pretty quickly in an intense storm. I’ve considered building a hybrid system that uses the faster comparator circuit with a high pass filter along with a micro to do some other tricks, such as holding the shutter open and then closing it after a strike or set amount of time for over exposure control. But I never got around to it due to the fact my PIC circuit works so well.