High speed X-ray video: jumping beans, wind-up toys and more!
Вставка
- Опубліковано 26 чер 2024
- High-speed X-ray video captured with a Dectris photon-counting detector. I show how the process works and how this detector is different than normal camera detectors.
www.dectris.com/
media.dectris.com/Technical_S...
Mexican jumping beans: www.amazingbeans.com/
Geiger counter: mightyohm.com/blog/products/g...
X-ray timelapse video: • X-ray timelapse of flu...
ImageJ image format converter: imagej.nih.gov/ij/ (the NIH's SSL cert expired?)
Flipping through images fast enough as if playing video: www.irfanview.com/
Video editing software: www.blackmagicdesign.com/prod...
The sequence of tiff files directly from the sensor contain a lot of temporal flicker -- probably because the X-ray tube itself has time-varying output. This isn't so bad at 60Hz, but quite a problem at 300Hz. I used Resolve's "color stabilizer" to maintain constant levels throughout a clip, and was impressed how well this removed the flicker.
Support Applied Science on Patreon: / appliedscience - Наука та технологія
Someone on twitter asked if there is a light-to-frequency conversion for Mexican jumping beans. Wouldn't it be funny to make a camera with jumping beans as pixels?!
I bet if you could articulate some kind of algorithm for light-to-frequency, had a medium large enough in scale to average out the imperfections, and fed an AI off of all that - the generated images would be fantastic.
Or what about the spectral response of jumping beans?
This is mind blowing, such setup in your lab
Also about uncompressed tiffs and an overall size and speed of output: as far as I know, substantial amount of cost of LHC, despite the complexity and high cost of the structure itself and its sensors, is actually a computational setup (I mean servers, network and so on), because there is no way to store information at that bitrate and enormous amounts for later processing, so data should be processed on the fly, and actually store only certain events with given conditions
@@hirobian2 so a smol lobster?
That would be fun! Yeas!
Awesome video!
Yeah it's cool! Hi Destin
Got the notification for this at the same time as a tinder match… i choose this every time.
Since when do science nerds get tinder matches
Ah yes, a man/woman of class.
Tinder match is more time-dependent, video likely isn’t going anywhere anytime soon. 🤷🏻♂️
@@dannymac6368 hes making a really trash joke nerd.
@@dannymac6368 if that tinder wonder go away under 15 minutes, maybe it's not worth for list "my future children's parent".. If you want to just took your toy to out have some fun, no tinder needed.
😱
Re. jumping beans and light, it might be fun to project a symbol on the floor with beans scattered randomly, and do a timelapse of them clustering around the dark areas.
That is a good idea
Imagine how often this exact thing happens in nature - too slow for us to grasp how cool it is. Kind of a natural sun-dial with alarm sound
Get the LEDs to go in a grid, have the beans move, then off all LEDs and take a picture. Stop motion 2D video using jumping beans!
Sounds like a good Cody's Lab title intro. ua-cam.com/users/theCodyReeder
@@BobCat0 I was thinking that too, totally something Cody would do.
I've worked at particle accelerator x-ray sources and we indeed had a bunch of these high speed photon counting detectors from Dectris and other brands around, though I've never seen such a small one. The Megapixel 2000 fps detectors I know of were in the lower six digits in price. The cool kids nowadays have Megapixel Megahertz detectors at free electron lasers though. Those capture short bursts of images at 5 million fps in full multi Megapixel size with large dynamic range
Wish I was a member of the cool kids :/
POV: you work at black mesa
I would imagine the large aperture ones are in the seven figures.
@@CircaSriYak That place really turned out to be a black mess, huh?
@@douro20 I think not quite. At 7 digits you're getting into the range of inhouse development at the respective facilities
I was laughing at that 6 figure delicate machine on your work bench by the vise in your garage, propped up on a roll of tape. You know they like you when they let you play with their toys!
You're so cool. Honestly, truthfully, I love how you do these experiments and demos which are fascinating and then go in depth about all the relevant details. I always enjoy your videos. :) Cracking your knuckles would have been great to see on highspeed X-ray if you're willing to accept the dose. 😺 A second or 2 at most to crack one phalangeal joint and capture that video.
Or just move your hand normally
His ultra nerdness has wrapped around into ultra coolness.
@@jhoughjr1 That describes the channel in a nutshell. Especially stuff like his own electron microscope.
I have joins that crack on the regular, without provocation. I’d love to see what that looks like in this thing. My ankles are particularly loud for some reason, I’m only 32… Either way, awesome video as always sir. I love when I get the notification for this channel
@@csours I mean, that's pretty straightforward. A synovial hinge joint like in the knuckle just glides as it bends. It hinges open and closed, so all you'll see are the bones rotating about their cartilaginous connection which isn't very visible due to its density. Cracking the joint though...that's an interesting phenomenon. It's a rapid dissolved gas release from the synovial fluid. The bones move quite fast.
Honestly the jumping beans was one of the coolest things I have seen. Great content as always.
Up until now I thought that jumping beans are something out of the cartoons. Wow.
First time see this insects and how they live
First i was thinking it a toy, lol
You mean one of the most cruel things
@@sheerun I said what I meant and I meant what I said.
i google it, and there is worms inside it
In 1972, Field Emission Corporation had a high power pulsed x-ray machine that ran at 1000 frames per second. The setup used a Varian image intensifier tube and a high speed camera with rotating mirrors so the film never slowed. The demonstration showed a model airplane engine running. The actual use was to image pig brains under high G loads using a rocket sled and rapid deceleration. The data was used by the space program and military.
Other products allowed stop-action x-ray images. One machine was capable of producing a 2.5 MeV pulse of 4 ns duration. There were interlocks to the room because the scattered radiation dose from one pulse was lethal. The unit was used to x-ray bomb detonations in the dessert, x-ray aircraft engines, simulate atomic blasts in small spaces, radiation harden semiconductors, and other things.
sounds very similar to the DARHT facility at LANL to image the implosion hydrodynamics of nuclear weapons core surrogate plutonium pits.
Back in 2010 I was touring a few universities to see where I might want to go, and met a guy at UCSD (I think) who told me he'd gotten an assignment to take a photo of a large explosion without detecting any of the explosion's own light, and he'd used X-ray imaging to do it. I wonder if that was related to this?
Seeing those bugs in moving x-ray is really cool, I don't think I've seen anything like that before.
I doubt they will send you anything much bigger than this since even this small one is inherently expensive, but I hope to be proven wrong.
Could you use this to visualize the X-ray emissions from peeling packing tape off the roll? That might be neat!
🧿🕳🧿
I think the trouble with that is getting an image to form on the detector’s focal plane. There’s no easy way to make a lens for x-rays. I think the way Ben is able to form an image in this video is that the dental x-ray tube he’s using is approximately a point source of light. So, tape may make a flash of x-rays, but the x-rays won’t form an image of the tape forming x-rays, if that makes sense.
@@bjmcculloch If lenses are impractical, why not use mirrors?
He should be able to I hope he sees this
Prob not. The triboelectric scotch tape peeling method of x-ray generation only happens in vacuum where electrons have a large enough mean free path to accelerate to KeV energies needed to create x-rays via bremsstrahlung. You'd either need to put the whole detector in the vacuum chamber or build an x-ray transparent beryllium window into the vacuum chamber wall (very hard).
Harbor Freight sells ratchets that are made of composite instead of steel. I bet it would be pretty cool to see the mechanism of gears/pawls inside working without having to disassemble the tool.
As he stated, the object can't be too solid for the x-rays to pass through...
@@Quickened1 i just said the body was made of composite, aka - plastic.
@@jong2359 being plastic is not the problem, it's the density of the plastic. Being a ratchet, it must be VERY dense to hold up to the kinds of forces a ratchet has to endure. It's highly unlikely the x-rays will penetrate, but I guess there'd be no harm in trying...
@@Quickened1 Sure, the composite will produce SOME radiopacity, but you will still easily see the steel mechanisms underneath. It takes an AWFUL lot of plastic to render an exposure completely opaque.
It sounds like you need to make a tiny mouse sized treadmill for next time.
Also, I think capturing all of the different types of Cherry MX keyboard switches would be neat. There are animated split views of already, but IRL x-ray video would be very interesting, especially if you add the sound they make on top of the video.
So cool! Finally got an old Faxitron DX-50 (12cm^2 sensor) X-ray and am attempting to add motorized CT scanning. Learning so much from your vids as usual!
That's so cool!
Actually photon coounting is also used in medical imaging. New generation CT scans with photon counting are making it to clinical use. They allow lower doses of both x-rays and contrast agents, higher spatial resolution, sharper images (higher signal to noise ratio) and allow "native" spectral imaging with higher precision.
At least that's what the manufacturers are marketing, I've never got to use one yet.
It allows ultra low dose chest CT that are studied for lung cancer screening in smoker ;)
@@TheThinkEat that's awesome! Thanks for sharing.
Does anyone know what the hurdles were for these new low dosage ct scanners? At first blush, it seems like the technology has been around for a long time. Given the cost of the normal machines, I would guess naively that the marginal cost of photon counting wouldn't be huge and the benefit of low dosage and reduced contrast seems enormous.
@@pkomarek the question would be best answered by the manufacturers themselves but I think that as for any electronics-driven technology, the main limits were miniaturization and efficiency. For this kind of application we need hundreds of tiny and highly reliable detectors concentrated in a small area (a few sq. inches), which might have been too expensive until recently.
In my work I have seen the big brother of this Dectris detector at a national lab synchrotron. Used for X-ray crystallography, essential for molecule structural analysis in drug discovery. And yes, they are eye-wateringly expensive.
I am a Technical Sales Rep for Proto X-ray Diffraction, and have been doing both single crystal and powder XRD for over 18 years. We use these detectors on our build for both types of systems. They are great for measuring diffracted X-rays from everything from small molecule organics like polymers to minerals and more. We are using one of these to assist in X_ray Metrology in a custom build for a customer.
This is the only channel on UA-cam where you can say "this is not just any x-ray video, it's high speed x-ray video"
Imagine my surprise seeing you having a Pilatus detector. My work place is probably one of Dectris‘ best customers, we have two of the current high end 16MP cameras, an even more specialised one for harder radiation, and at least a dozen smaller and older detectors. These things are used at accelerators dedicated to generating X-rays, so-called synchrotron radiation sources. Also funny that you mention ImageJ, we use it every day. Btw, the Mylar window is mostly there to keep the dry nitrogen in. The sensor is cooled with Peltier elements to sub-ambient temperatures and would form condensation without it.
I’ll send this video to my colleagues for sure.
A few things to film that come to my mind:
- Water spray bottle in action
- Lens focusing/zooming
- Lighter piezo thingy
- Non-electronic alarm clock at an alarm time
- Screw being driven into wood (or other materials?)
- Mouse wheel (different mice?)
- Mechanical mouse movement (the one where a ball is rolling)
- Car horn
- Fire (does it even show?)
- Popcorn
- [stretch] CD player loading a CD / floppy drive doing the same / cassette loader doing the same (I was always fascinated by the amount of engineering that went into these)
- Gluggle jug, or anything else from Steve Mould's videos where they made 2d versions
And oh so many others :)
Please do keep on filming them, it's so cool!
Honestly, one of the best (and one of my favorite) UA-cam channels that has ever existed. The content makes my jaw draw and me gently slap my head saying - 'how does he keep finding/making such interesting things'. LOVE IT!
I always look forward to these!
10:13 i use JpegView (with the play slideshow/as a movie function). With a good PC behind it, it can render TIFFs that are quite high into a good movie. A few years back i scanned at max resolution with a big specialized scanner, a glass photography collection and i was using several scanners at once, so i'd use that function (automated a little) to check that all the TIFFs were in working order. It's a good program.
Always look forward to watching Ben's videos.
Apparently the technology was developed at PSI in Villigen, Switzerland and was first used in the CMS experiment at CERN. It uses a microarray of reverse-biased photodiodes which I believe are driven into the avalanche regime.
The documentation for this camera warns you not to connect the computer to the Internet...
I never heard about the jumping beans, the whole thing was new to me!
Glad to see you back, great video as usual and appreciate the effort put together to share your experiments
Fascinating theme. Can't wait for another video of this sorts! Oh and as programmer and Linux enthusiast I'm fascinated with the technical setup. I can't even imagine how hard it is to debug all of this ordeal
This is so awesome. This channel is always fascinating. I often rewatch videos several times.
Very interesting, thanks for showing this to us!
Really good video. Seeing how things work without taking them apart, and just looking what's inside objects or living things is always exciting. Big thumbs up!
Your videos are always the ones I look forward to the most.
I was actually thinking about you a few days ago and wondering what you were up to and when you'd make another video. Glad to see you're still uploading. This was awesome as always. Looking forward to the next one.
It's always great to see you post!
just incredible, as always.
I've been patiently waiting for this exact video
This is so cool, I hadn't seen X-ray video at all before and the high speed is incredible
Dectris MVP! love that they are letting you use this!
Always a great video!
Extremely cool. I'm so glad you've still got a lab and you're still doing stuff like this.
👍Always love how you describe some of the mundane tech challenges, 32b tiff, image processing, etc.
I love this channel so much.
This is one of the very few contents (let alone UA-cam channels) that I instantly hit Like before even watching the video.
Good job and best wishes from Syria ✌🏼
This whole video just blew my mind.
Once again you show something like no other!
Excellent video as always can’t wait to the next video
This was AWESOME!
Now try neutron imaging - it's like X-rays in reverse: hydrogen-rich is black, metal is white. Although finding a suitable source _may be a problem_ :D
I guess he'd have to make a fusor.
Neutron imaging. Now that's something I'd like to get my hands on. Well, not literally my hands.
And here I thought David Hahn had passed away in 2016.
And - forget your complexion after that too. (Look what it did to Hahn.)
That’s insane! Keep up the great work!
Look at this maniac putting the kids toy popper UPSIDE DOWN.
Really cool video. Thanks for making it!
Your videos are so unique and enjoyable to watch - thank you 👌🏻
I paused as you went along and did the steps myself and that helped a TON.
Awesome bit of kit, there!
You always have the coolest projects.
Literally the coolest thing I've ever seen on this channel!
hm, he has a delorean
Awesome, as always.
I've been waiting for another video thank you sir
those contaminants on toys interest me alot. My dogs love tennis balls and things that have a cloth covering in general and ive always wondered what was in some of them considering their low cost and origin of manufacture.
Yea I was a bit surprised and concerned at his finding on that toy! 😵💫
Almost had to be metal fragments from the machinery used in production of the fabric.
You know they always write in ingredients section on food “may contain peanut traces” or something like that
I think there is a similar situation here
Manufacturers make a lot of stuff at the same factory, and a lot of stuff may end up in the glue or in molten plastic, for example metal chips/dust
Most of that is the plastic they use being recycled cheap stuff, because it is not visible and needs little strength compared to thickness. So they use the cheap contaminated scrap granules made from all the floor sweepings, pvc foil stamping with aluminium bonded to it (think of things like PVC blisters with a backing, what the blister was punched out of), old scrap lead acid battery casing that are ground up, and all the random plastic that gets sorted by colour alone. Mixed up, moulded into inside parts where not seen, and you often see marbling on them where different colours were not fully mixed
Perhaps the contamination, is actually there on purpose to help add balance to the toy
That was a fascinating vid! Thanks for having the idea to do this, it was quite interesting.
As ever Ben, you just one over everyone else. You work on another level!
Great as always
You should join the Safety Third podcast for an episode. I'm sure they'd love to have you on.
Thank you for explaining the jumping beans!
Very interesting and cool. Thank you!
The "this is youtube" part sounded more like "I don't want to get canceled" XD. As always, awesome video.
Amazing video again!
Honestly I get so excited when I see a new upload. You could literally just post any old video talking about anything mildy unique or interesting and we'd be watching it!
Okay now to watch this video lol
Wow those are some spirited larvae! I thought it would be cool to open a soda can but realized those were metal so it wouldn't work...then realized they sell in plastic bottles too 😂
Maybe a computer fan would be cool to see the bearings, or an old-fashioned metronome
I'd like to see a disposable lighter as the butane changes to gas at ignition port.
Super awesome! Let's get the big one!
Fascinating as usual
Awesome video and awesome device! So cool they loaned one to you.
I wonder if a light bulb filament moves enough when powered up to be interesting?
There used to be novelty versions that had a magnet in the bulb that would cause the filaments to wag back and forth... I'm sure to a degree there's an interaction with the Earth's magnetic field, make it sit there and shake
I did a video of a tungsten filament in slow mo . . . ua-cam.com/video/gshsoK4VOHo/v-deo.html
yes more please!! that is super cool!
Its been a long time, Ben! Nice to see you again!
Awesome! Fantastic content as always, thank you!
And thanks to Dectris for lending that instrument. Here's my request for them to lend you the largest sized one!
This is a very cool video! Thank you for sharing this Ben. I hope that you can get ahold of the super high-speed X-Ray detector.
Looking forward to the what else you come up with!
You actually x-rayed your finger in action. That's awesome man!
I kept cringing
@@jamescollier3 Meh. If there was a huge reason for concern, Ben would most certainly not do it. In the video, you could also clearly see the flesh of his finger, so they're low energy x-rays. They are x-rays none the less and all ionizing radiation is bad, but, I trust his jidgement.
There is also possibility to x-ray the "other" finger. I mean, for science...
@@skuzlebut82 Exposure time also plays a major factor. X-rays exposure is usually a fraction of a second.
@@skuzlebut82 I'm sure you are correct, but
Always such interesting content thank you .
amazing, absolutely amazing.
Nifty AF ! as always
Love the coffee mug. Also great to see more exciting videos. Always look forward to your videos 😁
Awesome video. YUM ❗. Thanks for great info to go along with the visuals.
Thanks for sharing Ben!
awesome
thanks to Dectris for sending this to Ben, send him more stuff please :)
you demonstrate a lot of interesting toys. thanks
A couple of years ago I wrote a tiff importer for Marmoset Toolbag 3, using libtiff. When you use that library, it quickly becomes apparent why tiff is sometimes referred to as "Thousands of Incompatible File Formats". Besides high bit depth images, its support also goes back all the way to 1 bit images, with various types of compression (or lack thereof), and file layouts. Programs often just seem to implement a few specific variations.
Yeah, TIFF is more a container for images and metadata and than a format. Not many people realize that even that container can have two byte orders and either 32 or 64 bit offsets. Then add ability to contain multiple images at different resolutions, possibly independently compressed stripes or tiles, compression can be LZW, Deflate, (less common is JPEG or basically anything) like 20 commonly found metadata records (tags) describing width, height, number of bytes per pixel, number of channels, meaning of channels, ....) and then some less common, but important for some applications. Not to mention that some programs are using these metadata creatively, so you can find SEM image with 254000dpi (example - 10000px/mm or 10px/um or 100nm/px - not even high resolution 2000px wide image shows 0,2mm).
And fun starts when you try to use TIFF for storing extremely large images in tiles - any program that opens it may try to read all data, display that image and either crashes or consumes a lot of memory. Fileformat itself allows it, problem is that most programs ... ehm ... none .... could handle some extreme use cases.
Incredible stuff!
Another cool video, thank you! I would crack my knuckle in the x-ray.
Jees always amazed about your knowledge ! its sick. really enjoyable !
Mr. Science, you're one luckiest guy in world of electronics and physics and you grow your hungry every month and suprised all of us. You are serious innovator of future Mr. Science juniors! You have make effect that never goes away, no matter if you stop everything today, BUT, PLEASE DON'T DO THAT!! Keep going and create ideas that anyone cannot even imagine. It isn't problem for you, and you enjoy everything all this. And get own counter that is in your pocket full day, and write down every day how much you get that dark power.. (:
Thanks man! You're God of technics.
Bloody fascinating! Nice HCT mug too
Super cool fun with x-rays. They are not as dangerous as people think - AS LONG AS the x-ray passes through the patient's body. I remember being worried about the overuse of x-rays in dentistry and orthodontics. Turns out I was worrying unnecessarily. In the early days, they wore these heavy, lead infused, aprons. These stopped the rays, but actually risked more damage.
Another amazing video!
I was just wondering when a new one would come out, thank you for this
Glad yt started bringing your vids to my homepage
Amazing. Total Recall IRL.
fascinating as always. thank you
Man, I'm always searching for this very thing... thanks!
Oh *wow.* I mean I figured you were up to something but this is a whole new level, hahaha. That jumping bean x-ray is ridiculous!