A light bending exercise… in space!

By Phil Plait | June 26, 2012 10:17 am

On board the International Space Station, ESA astronaut André Kuipers just put up this ridiculously cool and fun picture of himself playing with water in space:

Wheee! But what are you seeing?


Let me explain.

The really short version of this is that the water is acting like a lens, flipping his face over. But there are two images of André’s face in there, and one is upside down! What gives?

First we need to look a the drop itself. On Earth, sitting on a surface like a tabletop, water drops tend to be flattened. But in space, where gravity’s not an issue, water drops form little spheres. That’s because of surface tension, an imbalance in the electromagnetic forces between water molecules, and is a whole post all by itself! But for now, what you need to know is that in orbit where there’s no net effect from gravity, water droplets form little balls.

In this case, you can see the drop isn’t a perfect sphere; it’s big enough that it can oscillate like a spring, elongating in one direction. That’s cool, but doesn’t affect what’s going on here too much — it just elongates the image of his face seen in the drop a little bit.

But we’re not done! The astronauts injected an air bubble into the drop. On Earth, that bubble would rise and pop, but again, when gravity isn’t your master, the bubble stays put. So in the middle of the water drop is an evacuated sphere filled with air.

So what’s with the funhouse mirror stuff?

Ah, that’s because light can be bent! When a beam of light passes through water or some other transparent material, the direction it’s traveling changes, which is why a spoon sitting in a glass of water looks like someone bent it. This is called refraction, and depends on two different things: the material itself (different stuff bends light by different amounts) and the direction from which the light hits it.

The shape of the refracting material — the lens — also changes the image we see coming from the source. The curvature of the lens affects the direction the light is bent. In the case of light coming from outside a sphere of water, the light hitting the top of the drop gets bent down, light hitting the right bends left, and so on.

And one other thing: the path of light bends whenever it passes from one medium to another, so it bends if it’s going through air and then hits water, and it also bends if it’s going through water and hits air!

So now we can figure this all out. We see André’s face right side up in the background. As light coming from the top of his head hits the top of the drop, it gets bent down. Light from the bottom of his gets bent up. Light from the left side of his face is bent right, and from the right bent left. So in the water drop itself we see his face upside down, and switched left-to-right.

But that same light is also passing through the water into the air bubble. This reverses the reversal! For example: light from the top of his head hits the water and bends down. But some of that light hits the air bubble, and gets bent up. This counteracts the first affect, flipping the image rightside-up.

So the net effect is that light that comes from André and goes just through water gets flipped over and reversed. Light that goes through the water and the air bubble gets flipped again, fixing everything. Because the air bubble isn’t a perfect sphere, his face is a little distorted, but I hope you can — haha! Ha! — get the picture.

Now imagine that he could inject a second, tinier drop of water into the air bubble. We’d see a third, teenier image of his face, and flipped upside down again! And then an air bubble in that would flip it again, and then…

OK, I think you’ve got it.

My point in all this? Even something as simple as a drop of water gets complicated when you look at it carefully, and in a different environment. Also, the picture you have of the Universe around you may be bent in some ways that don’t always seem obvious at first glance.

When you look around you, how many lenses — both physical and mental — are distorting your view?

Image credits: ESA/NASA; Shutterstock (alexsalcedo)

Related Posts:

Squishy Moonrise seen from space
Mistical Moonrise time lapse
Ring around the Moon
Naked rainbows

CATEGORIZED UNDER: Cool stuff, Pretty pictures

Comments (42)

  1. Of course, we see this configuration on Earth all the time with soap bubbles; it’s just that the membrane of the bubble is so thin that there isn’t much distortion, or enough space to see the inverted image in the liquid portion.

    It would be pretty difficult to actually inject the drop of water into the middle of the air bubble – well, injecting the water is easy, but keeping it in the middle as you remove the injector would be hard, due to the surface tension causing the water to stick as you try to pull it out. Fortunately, that same surface tension means that water droplets will tend to bounce off the inside of the bubble rather than merge with the outer shell. (I have seen a video of that phenomenon although I’m unable to find it offhand right now.)

  2. tl;dr: there’s an air bubble inside a water bubble in space.

    I don’t know Phil, do you really think it’s necessary to explain that light can be bent? 50% of us wear glasses.

    And I think I have seen a marble before in my life. Or a crystal ball. Or a glass of carbonized water with air bubbles inside.

    I know this is a cool picture but… ah whatevs. Long live the longwinded explanation.

  3. MonkeyDeathcar

    So you weren’t interested in the explanation.

    I don’t know Sebastian, do you really think it’s necessary to explain that something was not interesting to you in 3 separate paragraphs? Some of us really don’t care what you think.

    And I think we’ve all read rants in comments that have no real purpose to them. Or in a newspaper. Or on the television.

    I know you think you have something important to say, but whatevs. Long live the longwinded rant.

  4. Dave

    Very cool!. ‘Nuff said.

  5. tardis42


    i found the video (i’ve seen it before as well)


    Don Pettit, on a previous ISS expedition (exp 6)

  6. JR

    I’m confused on one point: light entering the water droplet gets bent and inverted, but, before that light gets to my eye, it has to pass through the droplet again and into the air. Forgetting all about the air droplet for the moment, wouldn’t the reentry from water to air completely undo the inversion of the light, thereby rendering the inversion imperceptable? Clearly not, but, I need a MORE detailed explanation.



  7. Chris

    No doubt the video of him trying to blow a bubble of inside a buuble of water would have been neater.

  8. tardis42


    it bends inwards at both ends. look up “Lens_(optics)” on wikipedia.

    it has to do with the relative speeds of light in each material, iirc.

  9. HvP


    Just consider that the initial refraction of the light at the first surface is only half of the bending necessary to invert the image. The other half of the necessary refraction occurs at the outgoing transition layer.

  10. Randy A.

    Way cool picture!

    And thank you for the explanation! When I first saw the picture, I thought that perhaps it was two water drops aligned. But the water drop with a air bubble inside is:
    A) A better explanation for what’s seen in the photo;
    B) More cool!

  11. JR

    Aha! Thanks for clearing that up.

  12. Roach

    On first seeing the picture, I was also wondering if it was a bubble in a big drop of water or two drops in series.

    I wonder if two drops in series would look significantly different or not?

  13. BA, thanks for the explanation. It WAS what I thought it was when I first viewed the image, but to have it confirmed by one who twists light into enlightening ways is always good. 😉

    That said, it’s nice to know that astronauts aren’t CONSTANTLY tasked with things to do, hence have time to have fun and share that out of this world fun with we planet side. 😀
    At one time, it was considered worthy to fill every second of an astronaut’s waking moments with a task.

  14. JR

    Roach, based on the explanations provided by tardis42 and HvP, it would seem to me that “nested” lenses–like the air pocket within the water droplet–would produce different images than lenses in a series, i.e. you would only get two inverted images in a series. It seems to me, the difference is that light passes through only one half of the water lens (getting only halfway bent) before passing fully through the air pocket lens. Once through the air pocket, the water bends the light through its second half before exiting and then traveling to the observer. Please, correct me if I’m wrong.

  15. JR

    Addendum: It’s important to note that, after the light passes through the air pocket lens, two images are passed through the second half of the water lens to create the images we see.

  16. Marina Stern

    I’m so smart I knew what it was before I read the explanation.

  17. Chris

    @2 Sebastian
    do you really think it’s necessary to explain that light can be bent?
    Guessing that you are German based on the website linking to your name, you have to realize that Phil is an American and the educational system in the US (especially for science education) is not that good. So yes he does have to tell people that light can be bent because only by beating people over the head with these facts will it start to sink in.

    Actually Phil what would be nice for the others would be a little ray diagram to show how the light gets bent.

  18. @Tardis42

    Ah, I HAD found that one, I just didn’t realize that the free-standing wave experiment was part of the same video. I thought they were two separate ones.

  19. kat wagner

    What @Marina said, only with attitude. I knew it was an air bubble inside a water drop because if it were 2 water drops in a series, one would be out of focus. Trust me. I mean, look at the dude’s face in back – he’s all out of focus, but not in a bad way, heh heh!

  20. Timothy from Boulder

    The bending of the light as it passes through a sphere of water will not *necessarily* create an inverted image. It does so in this case — and in most cases — because the distance of the object (face) and image (camera) are such that the real image of the face is located very close to the droplet. The rays cross as they travel further from the droplet, producing an inverted virtual image of the face.

    If the object was *very* close to the droplet — say a droplet diameter or two — and the camera was similarly far away, you can get a right-side-up virtual image.

    If the object was *very, very* close — less than half a droplet diameter — it would be inside the focal length of the lens and you could be at any distance and the virtual image would be right-side-up.

    But for the typical distances involved with a face, a water drop, and a camera the explanation holds.

  21. #13 Wzrd1:
    “At one time, it was considered worthy to fill every second of an astronaut’s waking moments with a task.”

    Like on Apollo 7, when Wally Schirra and his crew objected to their excessive workload, leading to the saga of the “grumpy commander”, who refused to perform TV broadcasts, etc.
    IIRC, one of the Skylab crews also “went on strike” over their workload.

  22. tmac57

    Oh yeah ! Laugh it up guys! It’s all fun and games until someone shorts out a photon sensor!!!

  23. Infinite123Lifer

    The Universe is so trippy.

    Revealing secrets of the physical world seems to be reserved for a special few. For those who can, please do. Do it for the sake of wonder. Do it because time, space and matter have originated just the right set of circumstances to make it so, to give you a chance, to manifest an idea or imagine a possibility and test it and just maybe give you a tiny glimpse into what the physical realm has to reveal. For those who do attain a certain enlightenment about the physical world . . . much appreciated.

    That might be a bit much but when I stop and think about nature, like light, like light bending o_O like light . . . actually bending, does it really make sense to me? Eh, kinda uh no, not really yes no yes? Ah buckets, marbles, soda pop, crystal balls yes yes yes but what is REALLY going on?

    Harumpf and Hooray simultaneously for even being able to imagine some ridiculously complicated questions. To wonder and not know. To imagine and fall so short, to imagine the immensity of my shortcomings is to know what I cannot and trust and believe that the future holds growth.

    On the topic though:

    I think it is interesting that the air bubble in the picture looks incredibly spherical, especially compared to the water bubble. K, its rookie day here but I am surmising that is simply do to the fact that air is a gas and behaves as such and what is cool is that we can actually see the gas trying to escape in all directions contained within this heavier medium of water.

    Also, if ESA astronaut André Kuipers actual face is image #1, the reflected image of his face in the water is image #2 and the thence reflected image in the air bubble is image #3 and there is also boundary surface #1 where the outside air pushes on the water bubble and boundary surface #2 where the water bubble pushes on the air bubble then:

    How can I see an image IN AIR (image #3)? I am guessing because refraction depends on the material its passing through and air is just another medium, like water, just with less stuff in it, therefore the matter in the air, the molecules, are able to absorb and release the quanta or photons . . . hence a picture in a bubble of air? Or is the reflected light of Kuipers face appearing on boundary surface #2?

    If there were a vacuum in place of the air bubble would we still see a reflection or image #3? (assuming boundary layer #2 was possible to attain)

    Its still rookie day here but if images/reflections are due to matter absorbing and releasing photons than when light passes through a vacuum I am guessing there would be nothing to absorb and then emit the photons/light and thus no image #3.

    That question brought me to Snell’s Law, total internal reflection and the indices of refraction of different materials where n=1 in a vacuum and n>1 in a transparent substance, where I reluctantly resigned for the evening.

    btw not that its needed but com’on Sebastian what’s wrong with helping your fellow via altruistic endeavors. Perhaps its the only way to surpass mediocrity. After all if the Universe’s complete knowledge were 100 I wonder if humanity would even score a 1. Would 100 be so unattainable that we would actually be closer to the world of the canine or perhaps the amoeba in total understanding? Yeah, whatevs we can never know that, but longlive the longwinded wannabe

  24. Timothy from Boulder

    For those interested in the details of the ray paths, here’s a diagram showing

    1. Only rays going around the droplet (upright — rays on top stay on top)
    2. Only rays going through the water of the droplet (inverted — rays on top go to bottom)
    3. Only rays going through both water droplet and air bubble (upright — rays on top stay on top)
    4. Combined


  25. Chris

    @24. Infinite123Lifer

    If there were a vacuum in place of the air bubble would we still see a reflection or image #3? (assuming boundary layer #2 was possible to attain)

    Yes. The index of refraction of a vacuum is 1 (by definition). For air the value is 1.00029. For water is it 1.333. So air and vacuum are really close you wouldn’t notice any difference

    @25 Timothy

    Cool, what program did you use for that?

  26. AliCatZen

    Im wondering if density of the water helps as well as the rotation to keep the bubble in the centre?

    Also, I’m glad these things are explained, there are always people hearing it for the first time, like school children. It’s not always about you Sebastian. ;P

  27. Timothy from Boulder


    How can I see an image IN AIR (image #3)? I am guessing because refraction depends on the material its passing through and air is just another medium, like water, just with less stuff in it, therefore the matter in the air, the molecules, are able to absorb and release the quanta or photons . . . hence a picture in a bubble of air? Or is the reflected light of Kuipers face appearing on boundary surface #2?

    No. The image results from the droplet and bubble creating a lens that bends the light coming from the object (face). The location of an image depends on the curvatures and materials, and can be anywhere — behind the lens, inside the lens, in front of the lens. In this case, the rays appear to originate from a point inside the droplet close to the interface (see the 3rd diagram in my post #25). But that is just by chance; there is no reason it forms *on* the interface. And it has nothing to do with molecules absorbing and releasing photons.

    If there were a vacuum in place of the air bubble would we still see a reflection or image #3? (assuming boundary layer #2 was possible to attain)

    Reflection, no. Image, yes. Just think of it as a lens that you’re looking through.

    Its still rookie day here but if images/reflections are due to matter absorbing and releasing photons than when light passes through a vacuum I am guessing there would be nothing to absorb and then emit the photons/light and thus no image #3.

    No, that’s not what makes images or reflections.


    Standard optical design application.

  28. eyesoars

    You can do something a little like this on earth, too.

    You can create an ‘antibubble’, which is a spherical shell of air in water. They’re moderately stable, with just a bit of buoyancy, but not so much they can’t live for 30 s or more.

    The technique I used to use was a few drops of Dawn in a large glass of water. With a squirt bottle that gives a good, small needle-like discharge, suck up some of the detergent-water, and then squirt it back into the glass. If you’re careful and consistent in the flow rate, you can generate a blob of water that doesn’t wet and join the detergent water in the glass. If you end by gently squirting slightly more vigorously, you can push the anti-bubble down into the solution such that it will live for 30s or so before it floats up. Anti-bubble volume is typically an ml or three.

    If it pops spontaneously, you’ll get a bunch of little air bubbles.

    Dawn has changed its formulation somewhat, but it still seems to work OK.


  29. Infinite123Lifer

    Off topic:

    Thanks folks. In an effort to understand some of these concepts (without a background) I find myself getting confused very easily when the details arise. When I do understand something, in my experience, I find that there is always a deeper phenomena behind each phenomena to seemingly no end and when researching just a basic subject it is easy for me to get lost in the translations. Its no excuse for my ignorance.

    Despite my difficulty to conceptualize I do enjoy an immense passion for science. I also love to ask questions. With all regards to the blog I hope its ok to display that here. I love Bad Astronomy and it is much appreciated.

    I find without reflection I am oftentimes lost and to keep going back to concepts is the only way to sneak an inkling of understanding out of them. Takes awhile for some things to soak in. Wade through the material, recheck the definitions, reread what has been said, relax and let it sink in and hope this thick skull still has some absorption left in it. I even tried osmosis, slept with a How the World Works text for the longest time :) Some folks just were not meant to be top notch scientists, guess I am destined for something else but i will be damned if I wont keep trying to learn a couple things about my existence and the physical world around me along the way. Rather than inquire anymore I am going to slow it down a bit. Thank you for the time. Thankful rookie out.


    ps. I am not done with you sneaky wave-particle duality electromagnetic radiation.

  30. JR

    AliCatZen, sure, “like school children.” Or, perhaps, like Liberal Arts/Business people whose most recent optics lesson occurred in high school and explained the behavior of light passing through a concave lens, period.

    But, I agree, Ali: Can ego be an impediment to progress or what?

  31. Meg

    Okay, I want to see the ray-tracing diagram of this thing.

  32. Timothy from Boulder

    @32 Meg. Click on the link in my post #25.

  33. Starrtoons

    If I was in the astronomy business I’d be looking like crazy for a substance that can remain flexible, liquid (maybe that’s redundant), and transparent in vacuum — with the longterm goal of floating a giant blob of it somewhere out there in orbit.

    It would be a lens of unparalleled flexibility, and you could just let it float there and monitor the tiny visible distortions created in its transmission of light, as cosmic radiation or other “space wind” deforms the blob.

    Neat idea, eh?


    Nice article, but:

    The spoon picture what shows is the lens effect of a round glass: inside the water the spoon looks larger but not bent.

    The picture should be taken on a square receptacle or with the spoon just in the center.

    Thanks for the bubble article.

  35. It gets better ...

    The image is actually flipped twice more as you look at it – once by the lens of your eye, so that an upside-dowm version hits your optic nerve, then your brain flips it the right way up again. It’s why very small kids have difficulty reaching for stuff – their brains haven’t learnt to flip the image it receives yet.

  36. johnny

    Someone please answer it: Why isn’t the water droplet perfectly spherical? Why did it elongate? It should form a surface with the least surface area, right?


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!


See More

Collapse bottom bar