The reason photographic darkrooms are needed is because almost any amount of light can ruin the film or the photographic paper before they are fixed. Until then these things are generally kept in sealed, light-proof containers until they are ready to be developed. But there are a few things that can ruin film even then, most notably because some types of film are sensitive to ionizing radiation as well as light. This was famously how [Henri Becquerel] discovered that uranium is radioactive, but the same effect can be used to take pictures of cosmic rays.
In [Becquerel]’s case, a plate of photographic material was essentially contaminated from uranium by accident, even though the plate was in a completely dark area otherwise. Cosmic rays are similar to this type of radiation in that they are also ionizing and will penetrate various materials even in places we might otherwise think of as dark. For this artistic and scientific experiment, [Gabriel] set up a medium-format digital camera in a completely dark room and set it to take a 41-minute exposure. The results are fairly impressive and are similar to [Becquerel]’s experiment except that [Gabriel] expected to see something whereas the elder scientist was more surprised.
Like cosmic rays or radiation from uranium, there is a lot flying around that is invisible to the human eye but that can be seen with the right equipment and some effort. Although [Gabriel] is using a camera with a fairly large sensor that we might not all have access to, in theory this could work with more off-the-shelf digital photography equipment or even film cameras. A while ago we even saw a build that used UV to see other invisible phenomena like electrical arcing.
Is it cosmic rays though? Maybe he just photographed regular x-rays coming from thorium in glass lens or radon from air inside the camera. Without quantum spectrometer to see energy captured it’s impossible to tell. Also, if you leave eggs, frozen pizza or bannana on a photographic plate long enough they will also emit x-rays from potassium-40.
While interesting, l didn’t think this was ‘muonography’ as written in his article. His setup collected 30 second long integrations for a total of 41 minutes on a really nice camera.
The camera lens was fully covered, so he calls all of the points muon detections from outer space.
I strongly suspect this was more likely x-ray emissions from trace radioactivity of the common materials in our environment. All modern scientific imaging devices show similar patterns, and these are orientation insensitive. If it were cosmic rays (which create particle showers), the patterns should be orientation sensitive. I would also expect the temporal distribution to be highly non uniform, with few most of the time, and then a huge number occasionally.
Anyone interested in looking further into this might look at ‘cosmic ray correction’ algorithms for spectroscopy. Likewise, there are similar algorithms for astronomy. Most of the events captured are just a few neighboring pixels, but some events spill diagonally across dozens of pixels at much higher electron counts…
Similar correction algorithms are used in x-ray computed tomography image processing. A good cosmic ray hit can produce a nasty image artefact if not detected and removed.
Not saying that camera is not seeing muons, because it is: Some of those hits will be muons. But there are a lot of other processes that can produce noise like that.
There’s no experimental control. He doesn’t do a very good job convincing the audience (and prospective buyer of the ‘art’ he’s flogging) that most of those blobs are actually muon landings, and not something else.
Stick it down a mineshaft. Wrap in clean lead. Climb a mountain. Anything to change the muon count.
Detectors with suitable time resolution can at least use a veto detector to discard time intervals or acquisitions before stacking. Since there’s no easy way to shield against high energy particles, correlated hits in veto detector modules above (and below) the imaging detector can help reconstruct and classify such events. Similarly, one can use additional detector channels (however they may be implemented at even lower cost vs. performance level) to explicitly select short exposures for cosmic ray likelihood. But that’s probably beyond the scope of what [Gabriel] had in mind.
A bit of context regarding veto detector modules:
https://as.virginia.edu/news/what-heck-cosmic-ray-veto-detector
https://indico.cern.ch/event/981823/contributions/4295320/attachments/2253433/3823058/mu2e_crv_seminar_2021_05_tipp.pdf
All true and correct, but the OP uses 30 second integration times with dozens of expected (or purported) events per time window, so the veto/coincidence method isn’t very useful.
Au contraire – it’s very useful, but choosing 30sec isn’t. The camera seems to support continuous burst at 2-5fps and will be exposing with good duty cycle even with 1-2 sec exposures.
Maybe that’ll accentuate banding and heating of the sensor edges with control and A/D converter circuitry, but it’ll be easy enough to correlate those more granular exposures to veto data.
Thought that to be self-explanatory.
I would recommend graded-Z shielding as a better solution to reduce natural background radiation. 3-10 mm of lead, 1-2 mm of copper, 0.5-1 mm of cadmium would block far more natural background x-rays than an inch of lead alone.
In that order? And cadmium? Seriously? I’d love to see the modelling (or justification) for that one. Thermal neutrons are pretty scarce in that environment.
hmm. pity he did not show another image taken with the same amount of time passed. if the dots are on the same spot, the sensor has a fault on that spot.
That was my thought as well. It just looks like sensor noise.
What would unexposed high-speed film show under a microscope?
The “exposure time” would be months or years though – from the time of manufacture until the time of development. That might smear out and “hits” making it hard to tell a “hit” from a “miss.”
some of those hits may be what he’s looking for but there’s no way to prove it isn’t local x ray emissions or sensor noise, they didn’t even take dark frames? silly
Love the expertise in the comments. I also tried this with some success on the Raspberry Pi Camera Module v2. I think it worked as I had a stack of 4 cameras and took synchronized ~1s exposures with all of them and then looked for hits in all sensors that sit on a straight line. But I also had a ton of thermal noise and other junk that I had to fish the real events out of. Even then I’m not super sure these were cosmic muons.
More details (sadly paywalled) here: https://ieeexplore.ieee.org/document/8824441
But do message me at https://hackaday.io/mihai.cuciuc if you want a copy.
Tous les capteurs ont des pixels soit morts soit chauds… effectivement on les repère quand ils sont toujours au même endroit. Une pose de 40mn favorise grandement ce phénomène.
Sorry… in English:
All sensors have pixels that are either dead or hot… you can actually spot them when they are always in the same place. A 40-minute exposure greatly contributes to this phenomenon.
Some years ago there was a minor planet/asteroid search using a very large camera and very high gain. I think it was financed by Paul Allen. It used public examination of images to spot objects with motion relative to the background stars. In the training sets were examples of tracks caused by subatomic particles that were traveling nearly parallel to the CCDs and tracked through several pixels. They have a characteristic look with very sharp edges. I would expect to see these occasionally in images with this story. If there are none, do some larger sets and try to capture some as a reference. If they can’t be detected, then maybe a different explanation is needed.
“In [Becquerel]’s case, a plate of photographic material was essentially contaminated from uranium by accident, even though the plate was in a completely dark area otherwise.”
Henri Becquerel did not contaminate his photographic plate with uranium, not even accidentally.
In his research on fluorescence in some minerals after being exposed to sunlight, Becquerel began in early 1896 to investigate any possible connection between fluorescence and Wilhelm Röntgen’s recently (December, 1895) discovered X-rays, which, unlike visible light rays, penetrated opaque objects.
On February 26, 1896, Becquerel planned to expose some crystals of potassium uranyl sulfate to sunlight and then determined if the crystals emitted X-rays which would produce an image on a covered photographic plate. But the day was overcast so Becquerel stored the uranium crystals and the covered photographic plate in a cabinet. Days later he developed the plate, expecting to see nothing since the uranium crystals had not been exposed to sunlight. But, serendipitously, Becquerel discovered there was an image on the photographic plate, caused by the penetrating radiation Henri Becquerel called “uranium rays.”
In 1898, Becquerel’s doctoral student, Marie Curie, described the phenomenon with the term “radioactivity”.