Turn Your Camera Phone Into A Geiger Counter

Next time you’re waiting in the security line in an airport, why don’t you pull out your smartphone and count all the radiation being emitted by those body scanners and x-rays? There’s an app for that, courtesy of Mr. [Rolf-Dieter Klein].

The app works by blocking all the light coming into a phone’s camera sensor with a piece of tape or plastic. Because high energy radiation will cause artifacts on the CMOS camera sensor inside the phone, radiation will be captured as tiny specks of white light. The title picture for this post was taken from a camera phone at the Helmholtz Research Center in Munich being bathed in 10 Sieverts per hour of Gamma radiation from the decay of Cesium-137.

We have to note that blips of ‘bad data’ from a CMOS camera sensor aren’t unusual. These can come from electrical weirdness in the sensor itself or even the heat from the battery. [Rolf]’s app takes a reading of the noise floor and subtracts it from the counter. Radioactive decay resulting in Beta particles such as the Potassium-40 in bananas or the Uranium in granite counter tops don’t really register, although [Rolf] did have some success with Potassium chloride and a long measurement time. Still though, it’s a really cool way to turn a phone into a tricorder.



96 thoughts on “Turn Your Camera Phone Into A Geiger Counter

  1. You sure about that? If you are in a 10 Sv area you’ve got big problems (in other words, you probably are not replicating this experiment at home). Now 10 mSv is high but believable.

    1. He took it to the Helmholtz Research Center in Munich, set it up in front of a Ceisium-137 radiation source, and made sure not to be anywhere near the room while it was being exposed.

    2. It was done a the Helmholz Center Munich — the largest source at the so called Buchler device has10Sv/h at the opening of the caroussell where the source were lifted up – nothing for home. This is in a secured area with restricted access — Other sources at the area are up to 2000 Gy/h (e.g. Sv/h) . The up was tested for up to 2000 Gy/h at the PTB with pulsed radiation sources (X-ray) used for Angiographie in the medical area for example. With very short pulsed of course. And it still works well.

    1. I remember a guy who took a cheap webcam CCD and glued it to a smoke alarm radioactive element. With some software, he turned it into a very very random number generator and/or to seed and entropy pool.

  2. How can he be in a 10 Sv area?

    1 as Vince points out that’s an extremely high dose, as in dip your balls in Chernobyl dose.

    2 is a dose equivalent unit, i.e. it takes biological damage into account so dose a given radiation dose will be more Sv if absorbed by said balls than if absorbed by your toe nail.

    1. Nuclear research centers have very high powered radiation sources. I don’t think the guy was standing behind the camera zooming it in or out, that looks like it was done in post production.

      Works like this:
      place camera near aim-point of source
      leave room, lock door once certain no one is inside
      go to control panel
      select radioactive element to raise
      wait for alarms to go off
      listen to machine raise source from vault
      listen for machine to lower source back into vault
      wait for alarms to stop
      go unlock door, and retrieve camera-phone

      1. we taped with 1920×1080, the zoom I did in postproduction of course. Nowone is allowed to stay at the place (nobody wants it I assume). With this source you are 50% dead in 30 minutes standing at the camera position.
        The source is so strang that even behind the door there is around 10 µSv/h due to the reflected radiation. I wait in the room behind a thick wall until the red light is off (then the door unlocks).

    2. “Sv .. is a dose equivalent unit” For the Buchler device you can choose a calibrated distance for a given Radiation either the dose equivalent unit or energy dose. For the equivalent dose unit a so called water dummy is assumed behind the object simulating the scattering of the tissue (radiation gets back and sums up). When calibrating dosemeters the dose equivaltent unit calibration of the Buchler devices is used to get a H*10 calibration for example. You can also select a energy dose and leave out the water dummy (with different radiation resulting but not too much – the water dummy scatters around 30% back with our measurements). Then calibration is in Gy. Its easier to use but not very popular (we use it on our displays as a Gy display is more precise) – a diffcult topic and I am about to get deeper into it.

  3. I’ve had a couple of those Traser Glowrings (tritium gas exciting a phosphor material to make it glow) for over 10 years now, I forget when I actually bought them but they’re still glowing and I stuck them either side of a lightswitch near the back door so I can easily see where the switch is when coming home in the dark.

    I’ve read that tritium can be made relatively safe as its radiation doesn’t penetrate through thick enough material such as the thick acrylic(?) the Traser Glowrings are made from, but I’ve always been curious to know if any is leaking at all.

  4. I’m confused, why are people debating the facts about the experiment being in a 10sv area?

    Was this article edited earlier? The video clearly shows the experiment taking place at a research lab.

    1. It takes place in a secure place (research center) with high restrictions to enter (radioactivy Kontrollbereich), the sources are shielded when you enter the location in large lead (you can see the device containing the sources). The sources are used for professional dosemeter calibration and research experiments (http://www.helmholtz-muenchen.de/amsd/home/index.html) we work together with Prof. Dr. Hoeschen at the facility with also other projects.

    2. “I’m confused, why are people debating the facts about the experiment being in a 10sv area?”

      In my experience, there are a LOT of people out there attached to their cell phones 24/7.
      Likely the idea of putting the phone down and walking into another room without it is just incomprehensible, so naturally it would never occur to them as an option.

      Ever since I was a child, I always assumed that the mass proliferation of technology would enrich the lives of all of humanity.
      Sadly, it seems to have created a strange dependency on it instead, that I’m not sure is always a good thing.

    3. Honestly I thought 10 Sv would be enough to kill modern electronics. Apparently not. I know that 30 years ago when I was starting out and we measured a chip’s etched feature size with a ruler (just kidding) there was a PR frenzy that low dose xrays could damage your new osbourne-1 computer or whatever, so I assumed 10 Sv would irreparably lock up something much smaller internals, lower current, lower voltage, etc.

      It would be a very interesting research experiment to take common off the shelf microcontrollers, flash drives/chips, burner phones, and just see how much dosage they can take before soft errors and also hard failures start. Apparently the answer is a lot more than I expect…

      1. you need much much more to kill it, we tested x-rays up to 2000 Gy/h in a very short interval, the electronic survives. Its a matter of the dose. We soon do some testes with longer measurement times.
        In the Lap also Satelite equipment is tested with high doserates and you have very high energy particles in the outer space, for example in 11km hight we talk about 100 MeV myons, which you get on a flight even at this low height, but these are already converted particles. For the electronic in outer space you need some shielding at the right places. Also you will find some reports on degrading CCD chips in space (mercury mission for example), you start to see permanent damage, but the rest of the electronic still works (but it needs shielding).

  5. About the 10 Sv, he says in the video that it’s extremely dangerous and after half an hour of exposure, you have a 50% chance of surviving another week if treated immediately. That kind of radiation seems perfectly possible in a research environment.

  6. Creepy gif. The bolts above the opening made me think of eyes so I was freaked out for a moment, since it looked like a mario baddie was blowing snow at me. Well maybe not mario but the pic reminds me of something and I can’t place it.

      1. Why would the phone to be dangerous now? Ionizing radiation doesn’t do that.

        To the admins: the comments interface as viewed on my iPod requires me to hit the ‘Report comment’ button before the ‘reply’ button is visible.

      2. Exposing an object to non-ionising radiation does not make the object in itself radioactive. All the medical instruments used on you probably go through some form of non ionising radiation, yet they won’t retain a shred of that.

      3. @cmholm you will get neutrons on an airplane flight (>10km) and myons. The phone can detect them after they are converted by the plastic. we had one testflight with a comsic class 5 event for around 10 min recorded with some impacts (in 7km height) — the difference is you see lots of pixels switching on at the same time – big spots ! The event was proven by a satellite recording it also. But we have do do more tests. Airplane flights can damage the camera sensors indeed, the manufacturer of high precision sensors are aware of this problem.

    1. With 10 Sv/h we got around 3 defect pixels (not totally defect but noisy) in a Android phone we tested for 10 minutes in this radiation field.
      The Buchler Device (OB20) allows for maximum of around 10 Sv/h (e.g. 10 Gy/h) with the CS137 source – 5 Sv dose are 50% deadly within the next days. Some infos on this: http://www.helmholtz-muenchen.de/amsd/services/radiation-facilities/index.html There are source above 1000 Sv/h for irradiation.

    1. Yes after individual calibration thats possible we will have a professional line at hotray.com (our new company) also with dosemeters. Tests are done already at the german government institution for this. The phones have some interesting properties not available on conventional GM Tubes (I dont want to talk about this too early…).
      In general for this some steps are required as currently the variation in exposure time (automatic) prevents some phones to be used for low radiation values.

      1. @JB yes indeed, it was the SOEKs warning, but it always a matter of time (I don’t sleep on the crystal), with 10uSv/h you get a dose of 10 uSv in one hour for a minute or so its not a problem, but I always have to be careful and I try to reduce exposure. More critical is to handle pitchblend with the bare skin which tends to leav traces of the material on the skin. Also URANGLAS (look at ebay…!)- never drink out of it (intended to do so on some of them) Uran will react with water and gets into the body.
        Also don’nt sleep next to a radium watch (I will prepare some examples with old watches) – they are more critical.
        For the thorite you will find infos on Wiki etc. don’nt wear the crystal more than twice a year (its really intended to be worn as a gem !!!) YOu get such things in Thailand for example.
        But dont forget that a single x-ray can give you much more dose than this.

      2. @JB by the way : was not my finger …! But to be honest I had touched the crystal at a different time.

        I really love the Buchler device, there you never get in contact to sources. And even the background radiation in this facility is lower than around outside.

      3. @NoahFect here the requested video on the baseline measurement. I did a 60 min measurement, shown in timelaps (6000%) The iPhone here also showed the automatic BG calirbation done once after the initial noise determination. You can see some dots of the general noise hitting the camera also some smaller and larger impacts of the background radiation above the standard sensor noise.

  7. This is so fake. He had to put it under a 10 Sv source to demonstrate the effect. You would be dead in no time at 10 Sv. At natural radiation sources the contribution of radiation to the noise in the chip is much smaller than temperature caused noise. No way to measure through the naturally occuring temperature variations in a cellphone.

    1. As I said, the phones start at around 10 µSv/h to count. SOme better some less. Natural radiation is at around 0.1µSv/h. The iPad 2 can measure around 1µSv/h. The difference between noise and hits of radiation is that usually multiple pixels are hit and thats causing multiple counts in one frame. So even its more seldom due to the less area it can be distinguished from the noise. The temperature sensitiveity is there (you can see it but you will also see the measured temperature in the log file for compensation. There are usually two points reached. When charging a device (another info we get) the temperartur rises. And during operation is slighly rising. But the goal is not to measure natural background noise variations the goal is really to warn of radioactive contamined material or sources on higher rates. A thorium doted lens will have around 10-100µGy/h at the surface for example. Of course you can use a Geigercounter but its expensive also much more than the App costs.

  8. @steve: It qualifies as a hack because you’re using a smartphone’s camera as a radiation detector, i.e. for something it’s not supposed to be used as. Free has nothing to do with the definition of hack.

    1. In reality for most people a ‘hack’ (including those illegal deeds) means saving money, or this aspect is most important. In this case one buys software, instead of specialized hardware, and gets inferior results among with some convenience. Because of mediocre sensitivity of the hardware part, the whole approach is, and will remain, merely a toy. Toys, now, are sometimes worth the money due to the amusement. (Excluding the “I Am Rich” app, of course.) I would definitely pay for software that makes my phone (not some specific phone) able to detect background radiation and potassium fertilizer — but not until then.

  9. Perfect timing fellow geeks! With all the fracking and talk of uranium mining here in VA (US), I was just about to go buy all the parts to build my own Geiger counter. For $5 how can I go wrong! And the theory is sound. a CCD is perfect for picking up ionizing radiation. Although I don’t see how they can reliably detect alpha and beta particles through glass.

    ** AUTHOR: By default (on my epic) your app tries to use the front camera. Is the front camera “better” to use (for alpha/beta measurements) due to less glass/particle blockage?


    1. @Tweeks usually the front cam has less logic for pixel cancellation, and we get higher frame rates — alpha particles won’t go through the plastic (not even paper), beta particles usually are better on the frontcam indeed, the lense parts seem to be less thick till the chip. But low energy beta particles will be blocked (H3, C14, most of K40 but it usually has some gamma). SR 90 usually pass, so we had some phones where it was blocked. Radium (watches) or Uranium ore will work very good like pitchblend etc. We just prepare some new videos with examples.

    2. @Tweeks . my comment just disappeared – yes the frontcam is usually better (higher frame rate an less logic on pixel cancellation) also les plastic. ALpha particles wont go through — for beta particles those with high energy will psass, (H3, C14 won’t usually pass the plastic – for K40 only the gamma part which is 10%). SR90 works for most phones. Radium (watches) will work, Uranium ore also and of course CS137, and TH232 and some of the decay products.

    3. also important to know, that most activity also involves gamma radiation, except for PO210 for example AM241 also has gamma radiation, and most alpha emitters do. Beta radiation when hitting metal will cause so called bremsstrahlung (indeed the german word is used) which is x-ray and can be detected. Biut the percentage is low in comparison the the primary source and therefore longer integration times might be needed to detect them.

  10. I seem to remember something from my childhood about being able to generate gamma rays my simply energizing lead with high voltage. i think that’s sort of how xray machines work too no? seems like a much safe, more controlled way of producing gamma rays (than obtainingb/handling radioactive materials)… no?

    1. You can get official license exempt material legally from:

      Within USA thats easy shipped (USPS) and permitted for general public. For other countries you have to check before. If you want to get one I suggest SR90 as beta rays can be shielded completely. Much more safe than uncontrolled x-rays where you get skin burns etc.!! Really! But also read all savety information on handling radioactive materials etc. The sources are strong though permitted (a the surface 100 uGy/h for example) ! Important to know these are all point sources, to get a full body exposure you might a lot of them ! Don’nt sleep on these sources.

    1. DOnt wonder on the compass video the SOEKS geigercounter already shows 0.35 uSv/h when measuring the LED compass its due to the close position of the Ra226 compass (vintage Radium Watches/CLocks are similar in intensity by the way !) and for the H3 usually no radiation can be measured as its shielded by the glass

  11. Replacing the sensor glass with a pyrolytic graphite window makes any cheap B/W camera sensor sensitive to alphas, however the filters on colour cameras tend to shield most of them.

  12. Last year a friend wanted to sell a cabinet that was in his basement. He wished to post it on kijiji, so he also wanted a digital picture of it to upload. After the pictures were taken, a strange ‘speckling’ pattern of light (not seen when taking picture) appeared in the recorded image. After reading this article I believe the weird light pattern could have been radon gas radiation coming from the old house’s foundation somewhere. Has anyone else seen this effect ?

  13. Would be good to take a look at the picture. If there is a real high concentration of radon, that might be possible, and under low light condition. But I would ex cept only a few speckles otherwise its camera noise which could be large also (or JPEG artifacts on low light).

  14. Hi, the discussions re. my invention are on 4HV.org

    I’ve been working on this since 2011, and recently found a way to make it work with just the graphite and some other assorted parts with reasonable sensitivity.
    If someone would like to take this on commercially please PM me on the forum, or email andre atp lanoe dopt net
    (demunge and remove p’s)

  15. Hi

    I installed the software to Samsung A40 phone and tested it with several radiation sources (mean values at least for 10min):


    – Am-241 114 kBq; 51 CPM
    -Am-241 IAEA 353 kBq; 88 CPM
    -Am-241 from smoke alarm; 49CPM

    Values are quite low as americium is an alpha radiator

    Uranium glass bottle

    with phone software
    – 1-2 counts per 10 min -> makes no sence = noncompatible with A40 with alpha particles

    with a Mini-Monitor Series 800 with an alpha/beta sensor
    – no shield 1200-1500 CPM
    – with paper shield 900 CPM

    The cold comfort is that the Smart Geiger Pro probe is not working any better

    wishes, vmo

    1. Results after reinstalling and rebooting the phone x times -> total failure – finds no camera -> waste of money

      On the other hand Smart Geiger Pro starts working after reinstalling it

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.