Cinematography Mailing List - CML

Shooting Water Drops

Hi everyone,

Am working on a video installation for which I need to shoot water droplets as they are forming into shape and then follow them as they splash into water creating ripples. I don’t mind from what surface they'll drip or into where they're gonna fall.

My interest is to capture as closely as possible (full frame) the exact moment of their formation into shape, their fall and the ripples they create. Any suggestions in regards to the equipment and lighting? Is it possible to do it with digital cameras like a something that should be talked about in CML-video? Do I necessarily need to use macro lens? High shutter speed? Any ideas about the lighting? Thanks very much in advance.

Maria Pesli
Dir/Video artist


I presume you're talking about high speed stuff. There were several systems at NAB which shot high-speed video frames, including one capable of high resolutions suitable for filmout up to 1000fps and lower resolutions much faster. I'm not sure if this would be cheaper or easier than a high speed film camera, although it should be.

I've never been that blown away by interpolation solutions, especially from 25fps capture - it works OK to turn 60fps into 120, but it isn't nearly as good at turning 25fps into 50. At some point you're just tweening motion blur.

Phil Rhodes
Video camera/edit

On the subject of water drops - to achieve the shot as described sounds like a pretty tall order.

Tracking a falling water droplet whilst maintaining a full frame composition then coming to a dead standstill when it splashes would be near on impossible.

A solution might be to use a multiple still camera array with the top position occupied by a regular video camera and the splash position occupied by a hi-speed camera.

Tom Townend,

Tom Townend writes:

> On the subject of water drops - to achieve the shot as described >sounds like a pretty tall order.

It's not all that impossible, but it is difficult and requires proper preparation and the right equipment -- and some practice.

I would recommend some lab glassware to start with, namely a burette and an adjustable stand to start with. This way you can more easily adjust the drip rate and height of the falling drops. As long as there are no external forces affecting the drops, they will always follow the same path and form at the same point on that path.

As far as stopping the tilt is concerned, a mechanical stop fitted to a geared head can simplify that problem. Panning is obviously unnecessary.

The next problem is holding focus at macro sizes. This can really only be achieved with a lot of light or a great deal of practice. A strobe system like Unilux can keep things cool.

The simplest solution might be to film it in stages and put the sequence together in post. For instance, short tilts of the drop hitting the surface combined with a short tilts of the drop forming, etc. As long as the drop is seen to be entering the frame, it should cut together fairly smoothly.

While film would yield the highest quality image, high speed video has come a very long way in the past couple of years. The instant availability of a video image would be of enormous value is determining whether or not you got the shot, and what you might have to change to get it.

Check out Your local rep might be able to direct you to a system and people to operate it.

Brian Heller
IA 600 DP

Sam Wells writes:

> I would think shoot 35mm (Photosonics ?) so you could stay wider and >scan down the frame - just a thought, as I don't know how you'd do a tilt >here.

You probably won't get the frame rates you need from a 35mm Photosonics as opposed to 16 mm. Also, you'd run through a great deal of film.

Tilting to follow a drop is not that difficult, it just requires practice, practice, practice.

Brian Heller
IA 600 DP

Unilux strobe lighting (synced strobes) might be a big help also.... is there anything like that available in Greece ?

I would think shoot 35mm (Photosonics ?) so you could stay wider and scan down the frame - just a thought, as I don't know how you'd do a tilt here. Probably not the best solution in the scientific sense, but perhaps doable. (Otherwise, some kind of digital stitching between the drop forming & another drop hitting the water surface)

Sam "I feel out of my depth of field here" Wells

Seems the synced strobes are more useful than a very high frame rate???

Sam Wells

A couple things to consider :

- A Waterdrop takes roughly 300ms from fall to impact to rebound formation to column to decay. Go from there and your desired length of shot to figure out your frame rate. Usually they are filmed anywhere between 800 -1500 fps
as a rough guide.

- As far as the visual goes: Height of the drop and depth of the water it falls into is really important as far as what the drop will look like. Also adding a little soap to the water increases its viscosity and will make the drop more "beautiful"

- Photosonic has a working digital camera at decent resolution that can do those frame rates.

- Lighting :

Consider when a body of water is at it's most beautiful : at sunset. Build your own little "sky" behind and over the body of water, give it a little gradation and color to taste and refrain from using too much fill light from the front to build up contrast and definition in the ripples.

- Macro :

The size of the water drop and the ripples is bigger than you would think. Think coffee cup size if you want the ripples included. You will need a fat stop though to capture the event in decent depth.

- If you can't afford a high speed camera consider this :

Shoot a series of digital stills and build an animation sequence out of it.

Use a photogate trigger connected to a flash(es) that has a controllable incremental delay circuit built in. Drop goes through photogate, triggers flash a couple of milliseconds later. Increase delay by 2-3 ms every time.

Very important in this scenario: Make the body of water large (not deep) enough as to not raise its water level from the 120 or so drops that you are going to capture. Also you need a very controllable "drop gun" that releases the same size drop from the same height every time.

Hope that helps


Florian Stadler, D.P., L.A.

Many thanks to everyone who replied to my message - its been more than helpful reading your advices and I'll get back to you as soon as I have the first results. Not sure yet if I can get hold of the equipment some of you suggested but I got a general idea how to experiment with it.

I regards to the filters applied automatically to the messages, I'd like to say that I was oddly taken by surprise when I read my message being altered and couldn't figure out what happened. So, I would much appreciated it if these rules will be e-mailed to me to avoid similar confusions in the future.

Warm regards,

Maria Pesli

Copyright © CML. All rights reserved.