Water Droplets – Update 2

Some of you may have seen my original posting on the 18th January 2013 that reported on my initial electronic exploits with solenoid valves and transistors and the like.

You may also have seen my posting of 3rd February with its .‘One Royal Collision’ image.

This post is about the route I travelled between the 18th January and the 3rd February.

What I am striving to achieve is to photograph events that have become known as droplet collisions: water droplet collisions to some – but that would exclude those people who use other exotic liquids like milk! So I’ll stay with ‘droplet collisions’.

This entails allowing a single droplet of liquid to fall into a reservoir of liquid, allowing the droplet to bounce ( which it will do if dropped from a sufficient height ) and then to release a second droplet so that it collides with the rising column created by the bouncing first droplet. What will happen then is that the falling droplet will break over the column forming some amazing and unpredictable shapes. If that sounds complicated, have a look at my post of 3rd February for the hoped for result.

To see some other fantastic example of this, try searching for ‘droplet collision’ on http://www.flickr.com.

I started some months ago using a PhotoTrigger box and its laser-beam attachment. By this method you set the laser beam up across a reservoir and release a droplet from some sort of valve or improvised device suspended above the reservoir. ( I have seen a photo magazine article that suggests a plastic bag of water with a pinhole in it, then take lots and lots of pictures – and hope for the best! ) The PhotoTrigger device detects the first falling droplet as it breaks the laser beam, waits a length of time determined by you and then fires the flash guns/strobes just as the collision occurs.

My current setup uses a 12 volt solenoid-valve, a 100cc plastic syringe to feed water into the inlet side of the valve, a bowl of water situated about 20cm below the outlet spout of the valve, four strobe units ( more on them later ) and a Nikon D800 with a 105mm macro lens and 1.4x convertor – to get me just that little bit closer to the action.

You may have guessed that timing is everything in this! And should you venture along this road, much of your initial effort – once you have a viable system – will be to work out some basic timings. Note that the sequence of events we are interested in all happen in a very short space of time – we’re talking thousandths of a second here!

There is little point in quoting real numbers here – I doubt if they will match your needs exactly – but I’ll do it anyway. I’ll ignore the camera settings for now, but the basic timings we are interested in are as follows. First, how long does the valve have to remain open to provide the first droplet of liquid? ( Try around 80 ms[milliseconds] – if that time is too short a droplet will never form; if it’s too long you will get a dribble of liquid. ) Second, the delay between the two droplets ( try 30 ms as a start ). Third, how long the valve remains open to produce the second droplet ( see above ); and fourth, the time to wait before firing the strobes ( around 200 ms for a valve that is 20cm above the reservoir ).

And the camera?

The images are normally captured in a fairly dark room – although pitch black is not essential. The camera is set to a long shutter speed: the BULB setting, if you have it ( this holds the shutter open for as long as you have your finger on the shutter button ) or a shutter speed of 1 or 2 seconds. What is also essential is a good sturdy tripod and what is advisable is some form of cable release ( I actually use a wireless release system so that I do not have to touch the camera ), and I would also advise, if possible, using a camera with a mirror lock up facility: one touch of the shutter button lifts the mirror, the second touch of the shutter button actually takes the image. What we are looking for is to minimise as far as possible any movement of the camera.

To provide enough light I have been using four strobe units – others use more, some people up to seven or eight. Strobes that have a manual setting are recommended because what you want to do ( strangely ) is to turn the units down to around 1/64 of their total power. This is because, with strobe lights, reducing the power setting effectively shortens the duration of the flash.

For example, I have a selection of strobes, but my favourite is a Nikon SB800. At 1/64 power the flash duration is 1/32,300 of a sec. In case you think that’s a typo, I’ll repeat it: 1/32,300 of a sec. At 1/128 power the duration is approximately 1/41,600.

And it is that very short duration flash which catches the image for you. It is not just the camera because not many cameras could provide a shutter speed fast enough to freeze the action in these events. And it also explains why the darkened room is advised, but why you also want to use as short a shutter speed as possible in order to eliminate any stray light getting into the image. Any light that lingers too long will record on the final image as a movement of the water.

Well, I think that’s enough for now. I’ll follow this up with some thoughts on why I went down the Arduino route and how that has developed into a working system – but a system which, as I write, is still under development.

Thanks for reading.


2 thoughts on “Water Droplets – Update 2

  1. Hi, Calee
    Thanks for your interest in this. It’s much appreciated.
    And thanks for the link to the Cannon commercial. Some great photography there!
    You’re right – I liked it. 🙂

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s