If you want to skip ahead to see the images, you can find the imgur gallery here. Otherwise, read below for the details!


I've been trying to collaborate on more projects with friends recently. When I approached my friend Travis about working together on something he suggested we take pictures of soap bubbles. We'd both seen vivid pictures of soap bubbles against a black backdrop and hoped to do something similar. After a few photo and brainstorming sessions, we ended up with a repeatable setup to capture the kinds of images we wanted. In this post I hope to pass on some of the things we learned and share some of the resulting images.

Photo Session #1

During the initial discussions we drew some inspiration from this article from digital-photography-school. The color and contrast are excellent in those images and very similar to what we wanted to achieve. We also agreed early on that we wanted to capture a "film" of soap rather than a "bubble".

We learned from the article to keep the light source close to the bubbles. So for session #1, we tried to bounch flash bulbs against a white reflector as a light source:

After a couple of hours of experiments with light position we got almost no color from the images and were overall dissapointed with the results.

We programmed my robotic arm to dip a loop of wire into the liquid and re-position it in front of the camera. It worked great for repeatability but it was a little slow. I forgot to capture a video of it during the shooting, but here is a bonus video from earlier in the day when I was setting it up:

Photo Session #2:

After session #1, we did more research and put some time into preparation. First, we ditched the flash bulbs and replaced them with a softbox.

The softbox has a number of advantages for this type of photography:

  1. It is continuous. This makes it much easier to compose, focus, and time the photographs.
  2. It is a direct light source, so it is easier to get even converage than with a reflector (especially since we are so incredibly close to the subject)
  3. We suspect the spectral content of the softbox is more "broadband" by nature than the Xenon flash tubes. The light spectrum of a flash bulb appears to have strong peaks, which may mean the images ultimately have less variety in color. This is an area we'd like to experiment with more.

We also ditched the robotic arm and 3d-Printed a motorized contraption for reseting our soap film:

If you want to make your own, I uploaded the 3D files to Thingiverse. I am using some custom code from another project to drive the motor, but you can easily to do the same thing with an Arduino board and some 3D printer software.

The last thing we changed was to add a little bit of sugar to the soap mixture. We'd read that this can help produce longer-lasting bubbles and more vivid colors.

Finally, in session #2 wanted the ability to "paint" on the surface of the film. We knew enough from the first attempt that we would likely want to introduce some kind of swirl effect. For this, we used a small speaker driver with a plastic food dispenser pressed onto the back and sealed with plumbers putty. The driver we used is available from Parts Express. We ended up driving the speaker with an approx. 80 Hz signal and modulating the amplitude for the desired effect. It worked very well. This speaker driver has a unique design and is particularly well-suited for this application.

With those three changes, we immediately started getting better results in session #2:

After a couple hours of shooting we ended up with some very interesting and dynamic images:

Photo Session #3:

We were thrilled with the results from session #2. One of the only remaining issues was that quite a few of the photos had issues with sharpness – particularly at the edges of the images.  If you look at the setup in the pictures above, the plane of the soap film is tilted back away from the camera.  This is necessary in order to capture the color reflections from the soap, but it makes achieving focus much more difficult. If the camera was focused at the middle of the frame, the top and bottom of the frame would be out of focus. In order to achieve focus over the majority of the frame, we had to stop down to f/18 or higher. Even with bright studio lightning, this forced us to use a slow shutter speed and we lost some detail.

For session #3, we got around this limitation by renting a tilt-shift macro lens. For $30, we rented a TS-E 50mm Macro f/2.8 lens from Thursday->Monday. This is a great deal for a lens which has a $2200 MSRP. With the tilt shift lens, we were able to tilt the focus plane to match the plane of the soap film. This allowed us to open up the aperture and get brighter images without sacrificing focus range.

I 3D printed a small focus aid which could be placed on the frame of the bubble machine. The focus aid is designed so that the ridges on the front of the focus aid are at the same location as the soap film. The ideas is to place the focus aid on the frame. Use it to carefully focus, and then remove it and start the photography.

We experimented with including different shapes around the perimeter of the frame. It appears that sharp edges can help influence the direction of flow and change the characteristic of the images.

Here are some of the frames we tried:

Each one gave a different result and character to the images. Here is an example of the kind of effect the "ridges" on top have:

A "stream" appears to flow off each each of the spikes at the top of the frame. 
Also, the areas in between and around the border features often have some of the most interesting color and detail. 

As a final change to the setup, we continually added more sugar to the soap mixture throughout shooting. We used simple granulated sugar and mixed it in with a small metal whisk. As we added more sugar, we found the soap film lasted longer and longer.

We eventually found that if we added enough sugar and mixed up the film enough, we could get patterns to "set" into the surface and stay there for 10+ minutes. Here is a video of that happening (the vivid color starts to "set" in around the 1:20 mark):

With that approach, we ended up with sustained images which look like this. You can even "paint" certain shapes and patterns into the image based on where you hold speaker and for how long.

Here is a GIF showing the effect sped up:

And an image of the result from the video above:

With all of the added sugar (we added about 5 tbsp of sugar for 2 cups of water) the film becomes much more sturdy. After setting an image it doesn't "pop" and disppear like a normal bubble. Instead, it breaks suddenly then slowly recedes to one edge. This is what a sheet looks like shortly after breaking:

Results:

We ended up with hundreds of images. They are all so unique and interesting it is hard to pick out favorites. I am including a few here, then have put a bunch more in an imgur link for sharing.

There are more pictures at the imgur gallery here.

If you are just seeing my website for the first time, you might also be interesting in checking another photography project I did with water droplets here.


A huge thanks to Travis who co-worked with me on this project. You can find more of his photography, including some soap film photos, on his instragram profile.