How much space does a single photon occupy?

[Peter Mason]

If a photon leaves a source 4,2 light years away how far apart will it be from a similar photon it was adjacent to (say less than 10^3 wavelengths) when it departed. Does the inverse square law mean that individual photons get further apart and stay the same size or do they occupy a larger space (area). I assume relativity (special or general, I do not really understand) stops photons dispersing in three dimensions. Why? How do the photons line up in space behind each other? If we do not know how do we find out?

I then tried to make a single photon telescope but was deterred by a number of issues.

The stability of the tripod carrying the device. It would need to would maintain its position to better than 2,8 e-17 arcsec, for at lest a few seconds, just to look at the planet 4,2ly away. I suspect that the Earth wobbles more than this and certainly the telescope mount. I was aiming to look for an artificial source of protons considerably smaller than the size of the planet.

The smallest pin hole I found was 12,7 μmm that, if mounted at 43º gives a 857 nmm elipse on the detector. Thus the question how big is a photon? Can the photons from the distant source find a way through the pin hole?

As of the time writing the only SPAD device that I have found requires a helium compressor. I go back to the days when if we wanted a photo diode we used a germanium transistor and scraped the paint off. I was planing to buy an avalanche diode from RS.

I could take a broad approach and use the Earth's inherent gravitational wobble to scan larger parts of the universe with my avalanche diode, with the paint scraped off. A bit of Fourier to filter out the noise and still find the occasional photon streaming from now distant planets and having signatures just a bit different to the background but I need a few thousand of them.

My crude calculation shows the photon density from a 1000 Watt/m^2 source 4,2 light years away is small 1/10^20 m^2.

What does a photon do when it encounters a pin hole of ten wavelengths diameter after it has been isolated by the inverse square law and traveled 4.2 light years? I am encouraged by the thought that should I capture one photon I should see several behind it if I can get the quenching circuit to reset the diode fast enough. How far behind are they?

I am now looking for the slightest encouragement and help to refine the calculations before placing the order with RS.

 

[phinds]

You misunderstand the propagation of electromagnetic waves. During travel there IS no "photon", just a wave. Photons are the result of the wave interacting with another object.

 

[mfb]

Photons don't even have a position.

The stability of the tripod carrying the device. It would need to would maintain its position to better than 2,8 e-17 arcsec, for at lest a few seconds, just to look at the planet 4,2ly away.

You can look at stars at that distance with your naked eye, without any special alignment requirements.

A pinhole will lower your intensity and give a poor angular resolution. Where is the point?

 

[Peter Mason]

I was trying to isolate a single photon stream emanating from a 1 m^2 source 4,2 ly distant. The idea of the pin hole is to eliminate some of the extraneous clutter.

I now realize from phinds that there is a weak wave front extending over a large area which might be why it can be seen in an optical telescope. The question is how to isolate it the background of other interacting waves?

 

[Peter Mason]

You misunderstand the propagation of electromagnetic waves. During travel there IS no "photon", just a wave. Photons are the result of the wave interacting with another object.

Thank you for the reply. I knew when I wrote that there is a large and fundamental lack of knowledge on my part and was expecting to have a conversation about quantum effects. Your input has however provoked more questions when I consider how a photon arrives in my proposed detector.
As I understand it the wave must start from individual fundamental particles radiating excess energy they have in the form of an electromagnetic wave. Adjacent excited particles from a homogeneous source will have the same frequency but presumably not the same phase relationship. This I further suppose results in a wave that travels, without loss of energy that is another mystery for me, and interacts with other waves it encounters on the way here. This understanding is based principally from observation of sea swell that eventually dissipates allowing me out of harbour. There are often several wave patterns from different directions. Do photons contain directional information? I assume not. Is there a way of gaining directional information possibly using multiple detectors and mathematical combining them, thereby steering the detector more accurately and locking on to one wave source? Again given the sea swell example I am certain that with an array of pressure sensors it would be possible to determine amplitude and direction even if only given the FFT spectrum of each sensor and its position in the array. FFT the photon equivalent .
Does the rate of photon arrival contain information about the temperature of the source or does temperature simply change the wavelength and have no effect on wave amplitude. My assumption is that temperature changes both wavelength and photon rate. My interest is in the frequency and amplitude change of the photon rate.
I found a SPAD 32*32 array Peltier cooled but probably too big and too expensive.

 

[mfb]

Unless you place a single-photon source at the source, there is no "single photon stream". And even with such a source, you would only be able to capture a tiny fraction of the emitted photons. With a 1 m² source there is no way to collimate light well enough, not even if your goal is just to hit Earth.

Forget photons, that just leads to confusion (as seen in this thread). Have a look at waves and diffraction, that is all you need to understand telescopes.

We need telescopes with a diameter of meters to have a chance to see planets as individual spots next to their stars. Otherwise the image of the star is so large that the planet is lost in it.

Adjacent excited particles from a homogeneous source will have the same frequency but presumably not the same phase relationship

There is no such thing as a "frequency of excited particles". There are not even excited particles, only excited states of systems. And they do not have an associated frequency.

Is there a way of gaining directional information possibly using multiple detectors and mathematical combining them, thereby steering the detector more accurately and locking on to one wave source?

Only with radio waves or other sufficiently coherent sources where we can measure the phase. Combining them physically is done in interferometers, this can be done with visible light as well (see the Very Large Telescope for example).

Does the rate of photon arrival contain information about the temperature of the source or does temperature simply change the wavelength and have no effect on wave amplitude.

Hotter things emit more radiation, but temperature measurements look at the peak frequency, that does not depend on distance (unlike the intensity we receive here).

 

[sophiecentaur]

@Peter Mason
It's a good idea to start with the idea that EM energy travels in the form of waves and that photons are quanta of energy which only exist at the locations where the energy originates from or arrives at and interacts with systems of charge and mass. Any source of EM waves has a finite probability of being detected anywhere in space. Obviously, we can usually accurately predict where we will actually encounter a photon but. if you feel that you need to assign an actual size or extent of a photon then it would have to be infinite - if you have a chance of finding one anywhere - certainly somewhere within the beam you have formed. This is why it's a pointless exercise to try to consider the photon being anywhere until it actually triggers your sensor.
People who first come across the idea of single photon sources seem to think that means that photons can really be treated as little bullets but it would only be in experiments which involve such a sparse production of photons that one could actually identify an individual photon that's transmitted and received. The way the energy gets across the gap is still based on wave theory.

 

[Peter Mason]

Unless you place a single-photon source at the source, there is no "single photon stream". And even with such a source, you would only be able to capture a tiny fraction of the emitted photons. With a 1 m2 source there is no way to collimate light well enough, not even if your goal is just to hit Earth.

But there must be a continuous stream of waves otherwise they would have size and be able to avoid the sensor. Yes we would only capture a tiny fraction of the emitted photons but looking in the time domain they will have a signature.

I will look at waves and diffraction. I was avoiding the maths. Do you have any suggestion where to look?
I cannot control the source. That which I have in mind is an incandescent lamp driven from a 50Hz generator but this is just a prop any abnormal signature would suffice.
I appreciate the impossibility of focusing on such a small target at this range but I am encouraged by sophiecentaur's suggestion that there is a finite probability of finding the wave even here on earth. I was expecting to have to set up a lab on the reverse side of the moon:-). Yesterday evening I was watching waves break along the shore line, albeit on the television, Yes they were well formed having defracted from around the headland but it still seems to me to be a good analogy for the situation here. Is passing the wave through a pinhole¹ equivalent to the headland having filtered and polarized it first?

Somehow and admittedly with limited knowledge, the idea of looking further using a small sensor is appealing to me. Rather like probing a surface with an electron microscope. We can obviously capture something from a place known to be in the vicinity of a planet although there is an argument for looking somewhere else as there is a greater probability of something being there (sophiecentaur's infinite search).Does anyone have access to an optical bench and a single photon detector?
Thank you for the quality of this discussion. I tried this on another science forum but the discussion quickly got lost and was exclusively about the inhabitants of the planet.

It is an aside to the direction I would like this thread take but why is the wave unattenuated and somebody assure me that ALL this wave energy since the beginning of time does not account for dark matter/energy. Stephan Hawkin, the christian bible and probably others, as far as I remember, agree that in the beginning there was only light. (Peter)

1 Diffraction A wave exhibits diffraction when it encounters an obstacle that bends the wave or when it spreads after emerging from an opening. Diffraction effects are more pronounced when the size of the obstacle or opening is comparable to the wavelength of the wave.(Wikipedia)

 

[sophiecentaur]

I will look at waves and diffraction. I was avoiding the maths. Do you have any suggestion where to look?

It depends entirely on your level of Maths. To get anywhere beyond the very basics, you need some idea about Integral Calculus and how the infinitesimal parts of a wave front add together to produce a diffraction pattern.
Probably best to Google "Interference diffraction theory" terms and trawl through until you find something to your liking.

 

[mfb]

I will look at waves and diffraction. I was avoiding the maths. Do you have any suggestion where to look?

Optics books.

The challenge in exoplanet imaging is not the brightness of the exoplanets - many of them would be bright enough to be easily visible. The challenge is the nearby star, typically a billion times brighter. You need a very good angular resolution to separate them, and that means you need very large telescopes.
Resolving features on a planet would need telescopes with a size of many kilometers.

The detectors are not an issue, we have highly efficient sensors that capture nearly all the light that comes in.

 

[Peter Mason]

you need some idea about Integral Calculus and how the infinitesimal parts of a wave front add together to produce a diffraction pattern

I can do "some idea of" but it would end at solving a differential equation, which I have suggested in the past is something that distinguishes myself from our cat.
I was supposing that I would have a discrete time domain amplitude signal that I could auto correlate to characterise the source and presumably a "simple" FFT might yield something. However you suggest that the diffraction pattern will also be something to consider.

I could probably build or buy the bits to do the the first two by pressing a Raspberrypi into service and involving the local astronomy group (good for my French). The diffraction pattern analysis would be an order of magnitude higher cost and mathematical complexity, involving the array sensor I found in Italy. Although I suppose I could just look at the diffraction pattern or use the Raspberrypi camera to capture it. Looks like I have a project at least for proof of concept.

Can I look from the side?

The detectors are not an issue, we have highly efficient sensors that capture nearly all the light that comes in.

I agree coming to this I am surprised that single photon detectors exist and I am amazed that so little energy can travel so far and then still be detected. There must come a point where the wave front is so diluted by the inverse square law it is no longer detected.

 

[sophiecentaur]

I was supposing that I would have a discrete time domain amplitude signal that I could auto correlate to characterise the source and presumably a "simple" FFT might yield something.

Sorry butI have no idea what that means in the context of calculating a diffraction pattern or bringing a signal up out of the interfering signal from the star.

 

[mfb]

I agree coming to this I am surprised that single photon detectors exist and I am amazed that so little energy can travel so far and then still be detected. There must come a point where the wave front is so diluted by the inverse square law it is no longer detected.

With an 80 square meter telescope (the largest existing telescopes) and a sun-like star as far away as Andromeda (4 million light years), you still get multiple photons per second.
The background noise is important, of course, and you are limited by diffraction.

 

[sophiecentaur]

With an 80 square meter telescope (the largest existing telescopes) and a sun-like star as far away as Andromeda (4 million light years), you still get multiple photons per second.
The background noise is important, of course, and you are limited by diffraction.

I have not understood why the OP thinks it's a good thing to be dealing with just one photon at a time when any astronomical observation tries to maximise the number of photons it's using. I think there is a confusion between photons and resolution.

 

[Peter Mason]

Sorry butI have no idea what that means in the context of calculating a diffraction pattern or bringing a signal up out of the interfering signal from the star.

Neither do I, that is what I would like to find out.

Ok it may be useful to understand my background. In one sentence I worked as an electronic engineer making microprocessors blink LED's and what you do to make money though I simplify the situation we were quite innovative at the time. The nearest I get to this is measurement of the rate of Gama radiation with a photo-multiplier and a scintillator to measure the thickness of steel plate coming out of a rolling mill. The computer we used occupied a room in the basement and had a drum store of many killo possibly mega bits. Naturally my attempt here is to move things into an area that I can relate to a wiggly line on a cathode ray tube.

If we count the rate of arrival of photons dP/dt or discretely ΔP/Δt it will produce a nice wiggly line. The intention then is to point the contraption at the moon to measure the reflected signature of the only inhabited planet we know, then Mars and Venus for a look at those we suspect have no life. At that point I suspect it all gets cold boring expensive and frustrating gazing into a night sky preferably from the far side of the moon.

I was about to suggest that because light is part of the EM spectrum we would need to developer a new directional sensor similar to a terrestrial TV antenna but at 800 nm. Thinking this is a bit science fiction I searched google and found this. http://piers.org/piersproceedings/download.php?file=cGllcnMyMDE0R3Vhbmd6aG91fDRBXzEwXzIzMDYucGRmfDE0MDMyMDA0MzEwNA== at the second attempt, the first was a pay site.

With an 80 square meter telescope (the largest existing telescopes) and a sun-like star as far away as Andromeda (4 million light years), you still get multiple photons per second.

This is worrying How many photons per pixel per second and how many photons does the detector miss.

Now back to Earth a red telephone box needs paint.

 

[davenn]

and how many photons does the detector miss.

all the ones not coming in the front of the detector optics

 

[mfb]

I was about to suggest that because light is part of the EM spectrum we would need to developer a new directional sensor similar to a terrestrial TV antenna but at 800 nm.

Telescopes are doing exactly this. They are a directional sensor for visible light (or infrared, or UV, or whatever the telescope is measuring).

This is worrying How many photons per pixel per second and how many photons does the detector miss.

The fraction of light collected by the telescope (relative to the total emission) in my example is 4*10^-45. To capture more, you need a larger telescope, or be closer to the star.

 

[sophiecentaur]

Neither do I, that is what I would like to find out.

This sounds like a fishing expedition in your mum's washing up bowl. You don't know what you're after or bait and hook size. Don't be surprised when you don't get a satisfactory answer for your quest.
You discovered the term
"single photon source" and you want to find a problem to solve with one. There are zillions of websites and forums about astronomy (pro and amateur) and they discuss things like exoplanets on a regular basis. Read around an you will find the relevant factors in this branch of astronomy.
In fact, google the two terms exoplanet and single photon source. See how many hits you get (reputable sources) and that will give you an idea about the worth of the scheme.

The fraction of light collected by the telescope (relative to the total emission) in my example is 4*10-45. To capture more, you need a larger telescope, or be closer to the star.

. . . . . or take several extra years of measurements.

 

[mfb]

. . . . . or take several extra years of measurements.

That doesn't change the fraction.
It just increases the total light you get.

 

[sophiecentaur]

That doesn't change the fraction.
It just increases the total light you get.

It increases the signal to noise ratio. SNR imposes the final possible accuracy (discrimination) of all measurements.

 

[mfb]

Yes, but it doesn't change the ratio you quoted. A longer measurement time won't capture more than 4*10^-45 of the total light the star emits in this time.

 

[sophiecentaur]

Longer measurement time gives a narrower bandwidth. That's Shannon information theory.

 

[mfb]

This is completely unrelated to both my initial comment and your follow-up comment.

In visible light in astronomy, observation time is always more than sufficient to have a theoretical spectral resolution way better than the experimental limit. Increasing observation time "only" increases the photon count.

CODEX at the ELT will be one of the best spectrographs ever constructed, with R=135,000. The theoretical time it needs to achieve this? Not even a nanosecond. There is no astronomical observation that lasts only a nanosecond. Moving the telescope alone takes minutes.

This is a [B]-level thread. Please don't derail it with unrelated comments on advanced concepts that are not even correct.

 

[sophiecentaur]

This is completely unrelated to both my initial comment and your follow-up comment.

In visible light in astronomy, observation time is always more than sufficient to have a theoretical spectral resolution way better than the experimental limit. Increasing observation time "only" increases the photon count.

CODEX at the ELT will be one of the best spectrographs ever constructed, with R=135,000. The theoretical time it needs to achieve this? Not even a nanosecond. There is no astronomical observation that lasts only a nanosecond. Moving the telescope alone takes minutes.

This is a -level thread. Please don't derail it with unrelated comments on advanced concepts that are not even correct.

That is not a well informed post at all. Detecting a low level signal in the presence of a high level signal is only possible if the noise level is comparable with or lower than the level of the low signal or its effect on the big signal. Snr can be improved in several ways. You are effectively implying that imply that exoplanets should have been detected years ago, if all that counts is the level of light received from the star. We know that they have only been found in the last twenty or so years.
The time to complain that the level of the thread is too low for a lot of the comments was long ago. Size of a Photon can only be dealt with at the lowest level by a curt statement that it is a meaningless concept. Same I brought in Shannon because of your remarks about total number of photons from a star. As with any signal, it isn't the absolute level but the relative amount by which the signal is disturbed by what you are looking for.
I mentioned bandwidth as being relevant but, of course, noise figure and dynamic range are also important.
There are loads of google hits that will tell you about how Shannon affects measurement.

 

[Peter Mason]

The original post did get a short reply.

You misunderstand the propagation of electromagnetic waves. During travel there IS no "photon", just a wave. Photons are the result of the wave interacting with another object.

My understanding of how light works has been improved and I remain hopeful that a simple single proton detector can yield results. My research as you suggested yields the result that both FFT analysis and correlation techniques are used to analyse interference and diffraction patterns. Further the problem eluded to in my sketch is solved with a star shader. The maths is terrifying and is only resolvable with a clever algorithm that is also beyond me. There are plans to fly such systems in space since the distances needed are large with the 8m telescope they need to observe the planet. I have no idea how to quantify it but a pin hole is likely to be enough to make it achievable here on earth.

If a composite signal and noise are integrated over time the noise will reduce only if it has an mean value tending to zero and is then only useful if the signal of interest has non zero value W.R.T.

Shall I change the title of the thread. I to thought it has exceeded it usefulness. Any suggestion of exoplanet in the title is likely though to attract input about little green men. "Single photon long range camera" perhaps.

 

[Drakkith]

If we count the rate of arrival of photons dP/dt or discretely ΔP/Δt it will produce a nice wiggly line.

It will not. Camera sensors do not produce an AC signal that you can see on an oscilloscope. Not one that makes any sense at least. Most produce a digital signal that has to be processed in order to get the image, which consists of a number of pixels with a integer value for each pixel. The number represents, in some fashion, the number of photons detected. Unfortunately there are many, many sources of noise that also add to this value, drastically lowering the SNR.

This is worrying How many photons per pixel per second and how many photons does the detector miss.

Top of the line sensors have a peak quantum efficiency of upwards of 95%, meaning that they can detect nearly all of the photons striking the detector. This represents only a single wavelength though, and the efficiency across the rest of the visible spectrum is lower, but often still above 70-80%. You can see an example at this link. Just click on the "Back Illuminated" option and then select the chart at the bottom.

 

[PeterDonis]

Moderator's note: thread moved to the Quantum Physics forum, since the "photon" is a quantum concept.

 

[Paul Colby]

The smallest pin hole I found was 12,7 μmm that, if mounted at 43º gives a 857 nmm elipse on the detector. Thus the question how big is a photon? Can the photons from the distant source find a way through the pin hole?

Use a single mode fiber instead of a pin hole.

People have likely pointed this out already, but you're thinking about photons as objects like they are ball bearings or pieces of fluff. You can turn the problem around. Use the fiber as the source and a CCD as an array of detectors. If photon were objects like ball bearings, then one could argue that each time a CCD pixel detects a photon one can be assured that that photon didn't strike any other pixel. This property of hitting only one pixel and not any other implies there should be an anti-correlation between two pixel counts. The observed fluctuations in pixel counts should be (anti) correlated. If you do the experiment (I have) there is no correlation between pixel count fluctuations. Photons are not objects in the sense you are thinking.

 

[Peter Mason]

Camera sensors do not produce an AC signal that you can see on an oscilloscope

The proposal is for a single photon sensor probably an avalanche diode because I have not found a SPAD device that is not connected to a piece of specialised electronics and presumed to cost a lot. That signal I can see on an oscilloscope although counting individual photons to produce rate signal is probably the final destination. I Know that if I receive a photon at the detector it can be counted I was more concerned that the rate of photon collection from a remote planet would be too low to be meaningful. Is there any data on photon data rates from these objects? Answer in Hz/m^2 at the aperture of the device before the glass, mirror, pin hole, fibre, or what ever please :-)

Use a single mode fiber instead of a pin hole.

I will look at this.

photons as objects like they are ball bearings

But that is precisely how the literature describes the photon just after it shows line of electric charge sitting in a line holding hands like a flock of swallows on an overhead cable.

there is no correlation between pixel count fluctuations

Is that pixel count fluctuations on the same pixel? What was the source of photons?

Just for the record when I started this I had no agenda I was just curious although I had some vague idea of capturing a stream of photons from a wave front.
Is it possible that there is useful detail in the wave front at a single photon detector?. If it were possible to build a detector, optical tube, pin hole, sensor, electronics, ADC, for less that 99€ might it be an idea to point say 100 at Andromeda and use Big data to look at the result. (I never could resist looking too far ahead)

 

[sophiecentaur]

But that is precisely how the literature describes the photon just after it shows line of electric charge sitting in a line holding hands like a flock of swallows on an overhead cable.

I agree that's the idea one can gather from some of 'the literature'. But that sample is not from the informed section of Physics. The Corpuscular Theory of Light dies hard but we do know better, nowadays. (Einstein did not have that idea in mind when he showed the mechanism of Phtoelectricity).
To return to your idea. I can't be absolutely sure where your 'pinhole' would be placed. We have established that a wide aperture telescope is required for resolution so your pinhole would, presumably be in the plane of the sensor; perhaps just one pixel in size? My question is how to make sure the pinhole is in the right part of the telescope's image. Contrast is a massive problem and can be dealt with to some extent by masking the central maximum of the star's image. Assuming that's doing its best too help you, you have the problem of identifying a significantly different region in the image that moves relative to the star's position (over weeks or months). That should involve as much information as possible - the values of all the pixels that are available. Long exposure time allows integration (that's reducing the system bandwidth) and could allow identification of a region near the star as a possible extra, low level, source. Two or more time lapse pictures can tell you if it could be a planet (moving) - and not just a faint fixed star in the background.
As in all engineering, it's the actual numbers that count and it would be possible to estimate the size of planet and distance from the star that would enable the scattered light to be comparable to the diffraction spread of the star's image. This is where masking the star can be a great help, of course but the subtended angle of another solar system is pretty small and a challenge for even the biggest telescopes. You should do the calculations to give you a feasibility check at the first opportunity.
If you really want to explore this as a possibility, then why not get hold of some free telescope data from one of the on line sources. Find a star that's already been shown to have exoplanets and see if you can number crunch that data and dig the planet's image out of the shash round the star?

to point say 100 at Andromeda

From that comment, I conclude that you need to get a lot of basics sorted out for yourself.
Andromeda galaxy is very distant (millions of LY) and full of stars. I am not aware that anyone has had the remotest idea about finding exoplanets at that distance; it a whole new ball park in astronomy.

 

[Peter Mason]

Questions.
I have some (2) pdf files of papers that I have found that I think would be useful references that I would like to upload here. Is that legal? Possible? Desirable?

 

[davenn]

Questions.
I have some (2) pdf files of papers that I have found that I think would be useful references that I would like to upload here. Is that legal? Possible? Desirable?

are they peer reviewed papers ? if so, put them on some site and link to them :)

 

[mfb]

photons as objects like they are ball bearings

But that is precisely how the literature describes the photon just after it shows line of electric charge sitting in a line holding hands like a flock of swallows on an overhead cable.

That is how bad pop-science articles describe photons. This is not "the literature".

Modern cameras are sensitive to individual photons (where necessary).

Further the problem eluded to in my sketch is solved with a star shader.

For a 1 meter telescope and Proxima Centauri b as target, this star shade has to be at a distance of more than 5000 km (as rough estimate: (distance to star * telescope size)/(distance between star and planet)). Completely impossible on Earth, and with a space-based telescope you have to steer your shade around with a precision of centimeters, while keeping thousands of kilometers of distance to the telescope. Not impossible, but extremely challenging, especially if you want to observe objects in different sky directions.
With a pinhole you don't collect enough light to see anything interesting, and a 1 mm pinhole would still need 5 km distance.

See the example numbers I calculated in an earlier post. You need telescopes to get enough light.

Questions.
I have some (2) pdf files of papers that I have found that I think would be useful references that I would like to upload here. Is that legal? Possible? Desirable?

Link to them, or write the reference here if there is no proper online version available. Don't upload copyrighted material here.

 

[Peter Mason]

Andromeda galaxy

Oops. For Andromeda read Milkyway.

The Corpuscular Theory of Light dies hard

For a considerable part of my time on the planet I have been assuring people that I have been searching for the truth and when I tell them that the only reliable source of the truth is physics this seems to find resonance. Now you inform me that I have to choose a variation of the truth.

masking the central maximum of the star's image

This is theoretically highly mathematical
https://www.google.fr/url?sa=t&rct=...6/meta&usg=AFQjCNGUQJ-oo8sbxobLsh50Uc8pe_h78w

and why the petals are not a prime number, the answer is always a prime number, is probably more to do with engineering expediency than anything else.
I will also take a pragmatic view of the occulter position and probably use two pin holes, 1,0μm and 12,μm. The local astronomy group has a summer open evening next week. I will go and attempt not to frighten and confuse them.
How to point the arrangement, I have no idea what is achievable with current positioning systems. A question for my astronomer. The reviews talk of having the image nearly in the centre of view so probably not good enough, though I suppose repeatable would suffice.

 

[mfb]

With a 12µm pinhole you might get a few photons per second from the brightest stars. You will get a photon every few years from the brightest planets. And you have no way to distinguish this from all the background sources, even if you could collect data for years to get 1 or 2 photons.
With a 1µm pinhole you get a photon every few centuries.