| ||||||||
29 March (2024)
Fabrice Laussy reposted
@fplaussy Quando corpus morietur fac ut animae donetur paradisi gloria. Amen
25 March (2024)
18 March (2024)
Certainly, here is a possible line for your tweet.
17 March (2024)
Replying to @jfrossier
That is not compatible. As AI progress, more and more authoring/reviewing/publishing will be "not even read" by a human.
14 March (2024)
The pair correlation function of suspended spermatozoa and bacteria... soon in your optical field! (nature.com/articles/s4159…)
13 March (2024)
Also hired the great Vesto Slipher, who saw first everything that Hubble became famous for. Vesto could not observe Pluto himself but hired the guy who eventually did (still at Lowell obs.) Great place.
11 March (2024)
𝗧𝘄𝗼 𝗣𝗵𝗼𝘁𝗼𝗻𝘀 𝗘𝘃𝗲𝗿𝘆𝘄𝗵𝗲𝗿𝗲
Part 3: the circle.
Part 1 here:
Hello again, conspiracy physicists...
Part 3: the circle.
Part 1 here:
Hello again, conspiracy physicists...
Fabrice Laussy @fplaussy · 23 Feb 2024
𝗧𝘄𝗼 𝗣𝗵𝗼𝘁𝗼𝗻𝘀 𝗘𝘃𝗲𝗿𝘆𝘄𝗵𝗲𝗿𝗲
Part 2: the paper.
Part 1 here:
Part 2: the paper.
Part 1 here:
Here is the big secret finally revealed (if you didn't read the paper), of the nature of the circle that t ies photons of particular frequencies together, making them vanish from the landscape of all quantum possibilities: some photons do not want to be emitted together! Why not?
Remember the first "revelation" that two-photon observables exist independently from single-photon ones. Two-photon things happen that don't transpire in photoluminescence. Correlating spectral peaks means overlooking all that happens that you can only see with two-photon goggles
Here's a two-photon view of detuned resonance fluorescence. ϖ are frequencies wrt the laser (normalized to splitting), the color is the type of correlations photons come with (red bunching, blue antibunching). The laser is at ϖ=0, the emitter at +1, the ghost emitter at -1.
Resonance Fluorescence (RF) is always symmetric, even with detuning. The two-photon spectrum (2PS) on the other hand, is not. Note that little happens at (+1,+1) where the real emitter is! The emitter literally disappears from the 2PS. Two-photon physics is ruled by the ghost.
This is because we keep forgetting that two-photons exist at a different level than one-photons. If we rotate our axes and forget about the peaks, we can restore a better two-photon symmetry.
We can make it more obvious by highlighting (in dashed black) the regions of no correlations (g⁽²⁾(ϖ₁,ϖ₂)=0). Beside the lines that are symmetric to the bunching (red), there is a wobbly hoop that surround the circle of antibunching (blue). There's also another one on top.
So the two-photon symmetry is not perfect but it is much stronger than you'd think if you keep your obsession for spectral peaks. It is 𝑛𝑜𝑡 where the two-photon action takes place (only at the one-photon level).
Still, to understand our circle, it will be useful to remember what the Humboldt University of Berlin observed from the peaks, which we already discussed earlier on this platform:
Fabrice Laussy @fplaussy · 25 Apr 2023
Follows a thread on our recent paper «Two-photon correlations in detuned resonance fluorescence» via @ioppublishing
By filtering out the central peak (the Rayleigh scattered laser) @MaxSchemmer et al. saw that the photons coming from the remaining two peaks—split away thanks to detuning—are bunched (come in pairs). Weird because if you include the laser, they are antibunched (come 1 by 1).
Such correlations have a long previous history, e.g. @AtaUlhaqPK's nature.com/articles/nphot… paper a decade earlier, filtering the peaks but under high-driving (Mollow triplet). I want to keep filtering but I'll leave high-driving to another time. So we go back to low driving.
Hanschke et al. (et nos) had done something similar but at resonance and removing fluorescence, showing that antibunching g⁽²⁾(0) arises from 𝑖𝑛𝑡𝑒𝑟𝑓𝑒𝑟𝑒𝑛𝑐𝑒𝑠 between four terms 1+ℑ₀+ℑ₁+ℑ₂: coherence 1, bunching ℑ₀ & squeezing ℑ₂. Mess that up and g⁽²⁾≠0.
After the Berlin experiment, we revisited Lukas' interferences but for the split (detuned) case. We cross-correlated the peaks but also homodyned out the laser, seeing a beautiful transmutation of a messy caterpillar into a two-photon cascade butterfly (animation by Palomo).
Clearly these interferences of various quantum fields are important, as you can seize control of what they do. We have worked out a quite general situation of admixing coherent and squeezed states in journals.aps.org/pra/abstract/1… where we deal with single modes.
Now one can do the same with any number of modes. Take two: we can create and admix coherence and two-mode squeezing (either independently or jointly; coherence is always independent) and study its correlations. Tuning α, ξ and ζ we can two-mode antibunch, superbunch, etc. Cute.
But hey! when you detect two photons and keep their frequency through a physical photodetection, that's what you have: correlations between two "modes". Our sensor formalism describes physical parameters from modes which become, like the frequencies that they detect, continuous.
So it's not only subnatural linewidth antibunching that one can understand as interference of basic quantum fields, but all multiphoton correlations. At low coherent driving, there's a full theory of coherent states and n-mode squeezing admixture that remains to be worked out.
We did work out the two-photon detuned resonance fluorescence. The circle is where two-mode squeezing and a coherent state interfere destructively. If you remove the laser, lo and behold, the circle disappears, everything becomes bunched. It's like having Dalibard on steroids.
You can also interfere directly the squeezing and the coherent state on a cavity. The two-photon spectrum is pretty much the same as the two-level system, whose role is therefore (at the two-photon level) basically to provide mere squeezing, cf. Section 5 of arXiv:2402.14010.
There are some traces of non-Gaussian physics, but they are higher orders. At the lowest order, the spectrum is given by this beautiful, incredibly compact formula, a jewel in the two-photon crown:
What do we see here? Well, set the numerator to zero to get antibunching (blue) and you see an exact circle (rewrite as (ϖ₁+½)²+(ϖ₂+½)²=½). Now set the denominator to zero, and you see a triangle of divergences: the superbunching (red lines). Becomes finite to next order. Wow!
Why should we care? Because it's a fundamental thing: interferences and detection, the play of the wrathful and peaceful deities. Because this has implications. Because this has applications. Because it's the tip of an iceberg we have no idea how deep it extends into the sea.
Part 4 here:
Fabrice Laussy @fplaussy · 9 Sep 2024
𝗧𝘄𝗼 𝗣𝗵𝗼𝘁𝗼𝗻𝘀 𝗘𝘃𝗲𝗿𝘆𝘄𝗵𝗲𝗿𝗲
Part 4: The crème de la crème.
As promised earlier this year (part 1 here ), we collect the best snippets of our paper that the Royal Society published today. A fast way to read the paper:
Part 4: The crème de la crème.
As promised earlier this year (part 1 here ), we collect the best snippets of our paper that the Royal Society published today. A fast way to read the paper: