Schlieren Imaging: Seeing The Invisible

Have you ever seen an image like the ones below and thought, how the heck did they do that? If so then read on, because I will explain the physics behind this unique imaging method and how to do it yourself!

Imaging Optics

The basics of taking photos is that we have to measure something. That something is usually electrical values from a digital sensor. When photons of light hit the sensor they cause the sensor pixels to build up a charge. We can read these charges as a voltage and build up the different voltages in each sensor pixel into an image.

Lenses

To be able to image something we must first get the light to reach our sensor. We do this by using a lens that focuses the light to an imaging plane that we can place a camera sensor at. If we were to place an object between the light source and our sensor, we would see a shadow because the opaque object blocks the light from reaching the sensor. This means that all of the light outside the object gets passed on to the sensor and we see a clearly-defined shadow of the shape.

The Schlieren imaging method takes advantage of the air density changes in a medium to act as an additional lens. When light passes from one medium to another medium with a different refractive index, it changes direction. This is how the lenses are able to focus light. One way to change the refractive index besides using a different medium is to change the density of the existing medium. This density change changes the refractive index which then makes the denser medium act like a lens.

Denser mediums have a greater index of refraction, therefore light passing through cooler air would bend more than light passing through warmer air because the cooler air is more dense than the warmer air.

If we place a transparent object with different density (in our case we will be using cold air) we can see that it will bend the oncoming light. We can then set up our lens system so that if the light is bent even a little bit, it will not reach our camera and the area with the increased density will appear darker than its surroundings.

I had a 50mm glass lens with a 500mm focal length lying around so I decided to use that for this project.

Focusing

When an image is in focus it looks sharp and we can resolve fine details. When the image is out of focus them it looks fuzzy to us. This is because the light coming from a single point on our subject is reaching multiple points on our sensor (or our retina). This diffusion causes the light to expand and wash out the small details.

https://bpb-us-e1.wpmucdn.com/blogs.lcps.org/dist/b/2409/files/2019/04/new_-momen_lens_macro_compare_sample-1s9vr9c.jpg

To get a high-quality image we should use as small of a light source as possible, ideally a point source like an LED or fiber-optic cable. In my case, I used a simple phone camera light. It’s not perfect, but it works well enough.

I then set up my DSLR on a tripod and moved it around until the light source filled the entire lens. Then it was time to test it out

Recorded Data

After a few images to get the camera ISO, shutter speed, and exposure settings correct I filmed myself holding an ice cube between the lens and the light source.

Now if you look closely, you may be able to see wisps of darker air falling below the ice cube shadow. This is the region of colder denser air that is formed when the ice-cube absorbs the heat from the surrounding air. It’s pretty faint, so lets see if we can extract some more detail.

Enhancement

  • First we will add our video clip to the timeline
  • Next we duplicate our video clip
  • Next in the “Colors” tab we can add an invert color node to invert the duplicated video clip.
  • Finally we go back into the composite mode and select “Hard Mix” and set the Opacity to 50%

This result is far from perfect but it is able to give us a little more detail on the shadows from the denser cold air.

Improvements

All in all, this project wasn’t bad for a few hours of work but there are some improvements that we can make. One of which is we can write our own dedicated algorithm for image processing, this would give us much better results. The other thing that we can do is to place a knife-edge in the path of the light at the focal point. This makes the effect even more sensitive to smaller density changes because the light diffracted by the density gradient doesn’t have to change its angle that much to be removed from the image system.

https://www.researchgate.net/figure/Conventional-Schlieren-a-arrangement-of-the-knifeedge-b-typical-response-curve_fig7_10577880

I was unable to do this because due to the optic geometry and the camera lens that I was using, the focal point of the lens would end up inside the lens system that I was using for the camera. A third thing that we can do is to change the optic system and either use 2 lenses to make a “test area” where the light rays are parallel

https://www.researchgate.net/figure/Sketch-of-optics-and-light-paths-for-various-schlieren-setups_fig1_289724552

Another thing that is commonly done is to use a mirror instead of lenses. To do this the light sources is placed at twice the focal length of the mirror.

https://iopscience.iop.org/article/10.1088/1742-6596/1144/1/012097/pdf

Final Thoughts

Schlieren optical systems can become very complex very quickly, and I wanted to do this project to show people that you can get adequate results with a very simple setup. This is great if you wanted to quickly test something and see if it is worth creating a more complex, detailed image with a better system.