> Statistics of the electric field in THz TDS can provide much better results for layer localization at deeper layers (lower SNRs) than conventional deconvolution techniques.
Could this technique also be useful for crystallography, or electron microscopy?
Terahertz waves range in length from 1 micrometer to 1 millimeter. These are far too large to be able to image things at the scales of crystals. (That's part of why X-rays are used for crystallography - they provide higher resolution than visible light, because the wavelength is shorter.)
Petahertz electromagnetic radiation is what's currently used for diffraction crystallography; it's conventionally known as "X-rays" (or "gamma rays" if it comes from radioactive decay instead of from Bremsstrahlung.)
You want to measure the instantaneous amplitudes of a wave packet in the petahertz range at above the Nyquist frequency?
I don't want to say it's impossible, but I think you'd have to improve the capabilities of our instruments by about four orders of magnitude, which means you would have to invent several new major technologies.
Maybe if you digitize the time-domain pulses using some kind of controlled nuclear reaction? Electronics are far too slow (although I don't really understand why, so maybe that can be fixed). X-ray-stimulated nuclear reactions are going to be tricky to get to happen at all, much less to chain together into computational circuits, although in that case maybe you can use α particles or neutrons to carry the information around between the computational elements, simplifying the task somewhat.
It seems like, however you manage to measure it, the quantization of the radiation is going to pose really significant problems to measuring its amplitude accurately as a function of time, but maybe you can get around that by repeating the pulses a lot of times. But you'll need to make sure that you're emitting pulses of the same shape, rather than just pulses with a similar frequency spectrum and envelope but varying relative phases among the frequencies.
On the other hand, if you just want to measure the time-of-flight for X-ray wavepackets to sub-picosecond precision, without worrying about the time-domain waveform within the packet, I don't think that gets any harder or easier depending on the "carrier frequency" of the wavepacket — X-rays should be just as easy as terahertz light. You do need a few terahertz of bandwidth, so you aren't going to be able to do this with bursts of gigahertz microwaves.
X-rays' interaction with the material you're trying to image might be less or more convenient; their cross-section for backscattering is pretty small, but maybe you could scan the book from different angles and use total internal reflection to find interfaces between materials in the book with different refractive indices. This would also reduce your requirements for time-domain precision down into the hundreds of GHz range, which is a lot easier to deal with.
Subvocal microphones work (to some extent) by detecting the electrical impulses sent to speech muscles when somebody speaks mentally instead of physically. I wonder if this would work with involuntary recollection - I ask you to think of your password, and even though you won't divulge it to me, the signals for speaking it would leak to my electrodes.
Actually this is another example of Betteridge's law. If you replace "book" in the headline with "one letter flipbooks of 9 pages or less" (from the paper), then we'd have a counterexample.
Still really cool, but a long way to go before we can actually do what the headline teases.
The research has interests of museums that want to read books without opening and damaging them. Is there any chance the scans can damage them?
Also, what are the laws of the US government applied this to reading our letters. Our mail is protected by law from opening, but what if they aren't physically opened?
Yep, that is my exact train of thought. To me, examining email is against the spirit of the law - however, that is allowed because the law did not exactly forbid email (email didn't exist few hundred years back...)
IANAL but I believe it is illegal to read mail in this way. You have an expectation of privacy with a letter. Same with your house, you have a reasonable expectation of privacy within it. That being said, I believe if this is cheap enough they will start doing it in secret and the courts probably will not stop it. Whether or not they will attempt to use anything gained from it directly in court is debatable but it could lead to parallel construction cases.
That's not true. According to US interpretation of law they may scan every IMAP mail which was not deleted. So in their analogy any book in a public or cooperate museum may be scanned, there is no privacy protection.
Only in democratic countries you can expect some privacy.
Note that I didn't say "You have absolute privacy in postal mail." The two have different legal protections, and those on snail mail are incontrovertibly stronger.
You have to leave the mail sitting for a certain amount of time, at which point it falls under the "abandoned" part and can be picked up without a warrant. Unfortunately they didn't interpret email being read as being picked up, since it's still sitting in the server.
but since you don't have an expectation of privacy when a third party cloud service handles your data, according to the law, why would you have an expectation of privacy for letters handled by third parties... for example, the Government can legally read all your emails hosted by a third party if they are 6 months or older (http://www.businessinsider.com/when-can-the-government-read-...)
You don't need this technology to read through an envelope. You can get a spray (envelope X-ray spray) that makes the envelope temporarily transparent and then evaporates without a trace. The other way to read mail without opening it is to insert a shaft at the top of the envelope, wrap the contents around it, pull out the contents to read, and then reverse to put the letter back in the envelope.
I think I saw this stuff at the International Spy Museum in Washington DC; it was a much more interesting museum than I expected.
You don't need this technology to do it but it sure makes things easier and faster to do. Allowing you to automate the process and apply it on a huge scale, say for example every letter handled by the US PS.
Up until now these would have been specific and targeted attacks where the benefit was worth the effort of the task, with little effort required they will have no reason not to.
They use an ultra-short laser pulse which "gates" (enables/disables) the detector for a very short window of time. They repeat the stimulus+gated detection with different offsets in order to sample different parts of the signal.
The ultra-short laser pulse can be generated through purely-optical means. You have a laser pulse traveling around a cavity. Due to a combination of non-linear optical effects, each round-trip tends to shorten and concentrate the pulse. The resulting pulses can be much shorter than what's achievable with electronic generators.
Scan the shoe and know the optimal betting strategy? You'd need some pretty serious technology to hide your knowledge and come up with a betting strategy that's optimally undetectable as well.
I find this interesting because I've ever thought of it conceptually, the idea of a scanner that instantly scans closed books (instantly because one pulse of whatever EM wave you're using suffices). I thought about it after reading about the Timbuktu manuscripts that were saved from Boko Haram in Mali. I hear many of them are too delicate to open. I hope this works. It'd be awesome for so many use cases.
The sensing system used was from "Zomega Terahertz Corporation", which had a web site "z-thz.com". That company seems to have disappeared at the end of 2015, but it's still in archive.org. Can't find any info that indicates what happened to them.
Zomega was one of the few companies with a true terahertz system, at 2THz, or 0.006in wavelength. Other "terahertz" products seem to be around 100GHz, which is much easier to reach; there are 77GHz automotive radars.
Zomega systems had already been used to image hidden layers in paintings, so looking into a stack of paper wasn't a big stretch. This new paper is mostly about a data reduction technique.
48 comments
[ 3.2 ms ] story [ 105 ms ] threadCould this technique also be useful for crystallography, or electron microscopy?
I don't want to say it's impossible, but I think you'd have to improve the capabilities of our instruments by about four orders of magnitude, which means you would have to invent several new major technologies.
Maybe if you digitize the time-domain pulses using some kind of controlled nuclear reaction? Electronics are far too slow (although I don't really understand why, so maybe that can be fixed). X-ray-stimulated nuclear reactions are going to be tricky to get to happen at all, much less to chain together into computational circuits, although in that case maybe you can use α particles or neutrons to carry the information around between the computational elements, simplifying the task somewhat.
It seems like, however you manage to measure it, the quantization of the radiation is going to pose really significant problems to measuring its amplitude accurately as a function of time, but maybe you can get around that by repeating the pulses a lot of times. But you'll need to make sure that you're emitting pulses of the same shape, rather than just pulses with a similar frequency spectrum and envelope but varying relative phases among the frequencies.
On the other hand, if you just want to measure the time-of-flight for X-ray wavepackets to sub-picosecond precision, without worrying about the time-domain waveform within the packet, I don't think that gets any harder or easier depending on the "carrier frequency" of the wavepacket — X-rays should be just as easy as terahertz light. You do need a few terahertz of bandwidth, so you aren't going to be able to do this with bursts of gigahertz microwaves.
X-rays' interaction with the material you're trying to image might be less or more convenient; their cross-section for backscattering is pretty small, but maybe you could scan the book from different angles and use total internal reflection to find interfaces between materials in the book with different refractive indices. This would also reduce your requirements for time-domain precision down into the hundreds of GHz range, which is a lot easier to deal with.
Terahertz TDS already samples at frequencies beyond the capabilities of bare electronics (look at my other comment on this story).
All that's left is thought scanning and then we will think about rethinking passwords.
Still mostly science fiction, but plausible.
Still really cool, but a long way to go before we can actually do what the headline teases.
Also, what are the laws of the US government applied this to reading our letters. Our mail is protected by law from opening, but what if they aren't physically opened?
Only in democratic countries you can expect some privacy.
https://consumerist.com/2013/07/03/forget-the-nsas-hi-tech-s...
I think I saw this stuff at the International Spy Museum in Washington DC; it was a much more interesting museum than I expected.
[1] http://www.produktinfo.conrad.com/datenblaetter/600000-62499...
Up until now these would have been specific and targeted attacks where the benefit was worth the effort of the task, with little effort required they will have no reason not to.
The ultra-short laser pulse can be generated through purely-optical means. You have a laser pulse traveling around a cavity. Due to a combination of non-linear optical effects, each round-trip tends to shorten and concentrate the pulse. The resulting pulses can be much shorter than what's achievable with electronic generators.
Zomega was one of the few companies with a true terahertz system, at 2THz, or 0.006in wavelength. Other "terahertz" products seem to be around 100GHz, which is much easier to reach; there are 77GHz automotive radars.
Zomega systems had already been used to image hidden layers in paintings, so looking into a stack of paper wasn't a big stretch. This new paper is mostly about a data reduction technique.
Overview of terahertz technology.[2]
[1] https://web.archive.org/web/20151220072732/http://z-thz.com/ [2] http://spectrum.ieee.org/aerospace/military/the-truth-about-...