Aux sciences citoyens !

Impliquez-vous dans l'aventure scientifique de demain : remplissez notre formulaire d'adhésion et entrez de plein pied dans la "Science Citoyenne".

Hypertélescope LISE est une association d'intérêt général à caractère scientifique. Vos dons feront l'objet d'un reçu fiscal qui vous donnera droit à un crédit d’impôt de 66%.

Adhésion et don à l’association doivent faire l'objet de deux paiements séparés.

Voir l'onglet Participer ci-dessous

Presentation Astronomia

Presentation Astronomia – 2014

On the occasion of the seminar by 2014, the website has published an excellent presentation accurate and detailed of the hypertelescope project. A valuable document to understand everything on the hypertelescope. Are you the reproduced here.

Until today

The achievement of a larger telescope mirror has always posed serious problems. Pour a glass of large diameter, very homogeneous block, without bubbles, slowed down the construction of the 4 m class telescopes. To go beyond the Bolshoi's 6 m telescope, it was necessary to abandon stiffness afforded by the thickness, synonym of unbearable weight. Flexible mirrors maintained by computer brought a solution, allowing to reach 10 m. But beyond that, it faces another problem, which is that of transport. It is not possible to produce the mirror in place, and for the diameters, it becomes impossible to carry it. The solution was to make mirrors of several parties, but composite mirrors.

In a composite mirror, it minimizes the interstices between fragments, in order to lose the least amount of light possible. Obviously, this is a computer that manages mirrors, to ensure at any time that their juxtaposed surfaces form only a single surface, that would have the equivalent monolithic mirror. But it is difficult to build the very large mirrors, and also their cost would become quickly prohibitive. Telescopes whose construction is currently envisaged the ELT (Extremely Large Telescope), aim a size up to 40 m. But for such an instrument (EELT, European ELT), must be a thousand mirrors. This brings great complexity in the precise positioning of each along the theoretical sphere.

Another route was travelled, also compatible with the previous one: that of interferometers. An interferometer is a constituted instrument of several separate telescopes working together. They point the same object simultaneously, and the light beams that come out are mixed so as to produce interference. The treatment of the fringes by mathematical methods (deconvolution) then to rebuild the image of the object. The advantage of an interferometer is to obtain, with acceptable diameter telescopes (8.20 m for the VLTI for example), a separator power equivalent to that of a telescope with a diameter would be the distance between the telescopes. The major drawback of this technique is the smallness of the field.

A mirror dotted…

The concept of the hypertelescope is to achieve a mirror with holes! One can easily make an experience. Take the small mirror that sits on the bench of the bathroom. It has a concave magnifying face. If you turn to a sunny landscape, we can in projecting the image upside down on a screen. Cover the mirror with a sheet of paper in which holes are made. You can still see the image, although it is less luminous. Where the idea to cut the mirror into small pieces, and place the pieces as they were before cutting. That is, they must be the same overall spherical surface. Can be surprised that the holes do not see in the resulting image, but it is so.

A mirror, made up of the juxtaposition of fragments leaving between them a free space, is named diluted mirror. We can understand that it is easier, and especially much cheaper, to realize a diluted mirror than a full mirror. Of course, the brightness will be less important (it is proportional to the collector area = the sum of the surfaces of small mirrors), but the power splitter will correspond to the outer shell of the mirror. We can then imagine to build very large mirrors diluted, with a set of small mirrors, so easy to make and inexpensive.

This solution had been considered for a long time. But it seemed unfeasible due to diffraction. Look at the landscape through a muslin curtain, and you'll soon: the landscape is drowned in a flare of light produced by this diffraction. This finding led to a rule for the construction of an interferometer, which says that the exit pupil should be similar to the entrance pupil. However Antoine Labeyrie has found it is underinclusive. It has shown that it must be same layout, but not similar. In fact, light brushes from the various openings must keep their relative positioning, but their dimensions can be modified. If you enlarge them, it increases the brightness of the image, and decrease that of the halo.

The drawback of the method still resides in the diffraction. By increasing the surface of secondary pupils, decreases the constructive interference which produce the image, and thus reducing the field.

How to make it?

We can consider putting a small number of large mirrors, or a large number of small mirrors. What is the best solution? Antoine Labeyrie has done tests, first by computer simulations. And the result is net: it is much better to use a large number of small mirrors. In fact, the size of the mirrors is not critical; and it depends on the brightness. The important parameter is their number. Over there, better will be the quality of the image. Now, we'll see how the image is formed.


The light from a star forms a plane wave, because the star is so far an excellent approximation, the spherical waves coming in can be regarded as planes. The wave reached the center of the lens first, and enters the glass. There, its speed is much lower than in vacuum, it is limited to 200,000 km/s. This delays the wave. On the other hand, on the edges of the lens, glass thickness to cross is negligible, and it wave is therefore not delayed. This is what brings the curvature of the wave after the crossing of the glass.

instr10The image is a task of diffraction

Light from two stars arrived on the lens in the form of two systems of plane waves (figured in two colors).


The two tasks of diffraction are far depending on the angle of separation of the two stars. Separation is sufficient, it is actually two spots if they are confused.


The focus of the image is obtained by slowdown in parts of the wave, so as to produce a concentric wave. A mirror, by its shape, gets the same result.

Full mirror
Regular diluted mirror
Mirror reduced to 2 fragments

To the left, we see what would be the image (bottom), seen in a giant telescope. Let's replace the complete mirror by a set of smaller mirrors, spaced. The image would still be recognizable, but much complex. Note especially that the light is spread out and form secondary tasks. How can one recognize a star field, if each gives such an image?


On the right, you can see the image that would give the mirror if he stayed more than two elements! These are the Young interference fringes. They contain the same information but well hidden… It's as well that it built the interferometers.

F14 F24 F34 f44
Fizeau interference, with respectively 2, 3, 5 and 9 aperture images Antoine Labeyrie

The images obtained, on the basis of the number of mirrors used, show that the number increases the quality. It should be noted indeed that the Central, non-existent task with two openings, is more and more brilliant when you add. Which means that the light is more concentrated in the diffraction task, instead of spread in annexes figures, which are instead more and more dark.

Well, the first experience of this kind was made at the Observatoire de Marseille by Edouard Stéphan, on Foucault telescope. It concerned only two openings. It does not allow him to measure the diameter of the star, because it is too small, but he managed to set a maximum Terminal 0.157 ".

Principle of the hypertelescope

The hypertelescope, as it has been defined by Antone Labeyrie, is an interferometer multi-login with densified pupil. This definition specifies all of the components that comprise.

Diagram of a hypertelescope, according to Antoine Labeyrie

It is considered a diluted lens, whose beams converge toward the Fizeau home. If the target fragments are arranged regularly, the image is visible in the centre: a central task, with secondary peaks. After the home, place a lens that makes the afocal (collimated) beam. Then, for each small beam corresponding to a fragment of the objective, it uses a Galileo telescope upside: its eyepiece (turned to the arrival of light therefore) is diverging, while the lens is converging. The eyepiece spreads the beam, that goal makes again parallel (bezel set afocal device). Notice that now, there is little space between the bundles. That is why this device is named densifier of pupil. Finally, a last lens convergeante reform the overall image. Secondary peaks have been collected, the central peak is greatly amplified. The secondary peaks are more very unpleasant, and can for example be a double star, which we'll see the two components.

Note that the pupil densifier is, again, an invention of Antoine Labeyrie…

The process is less good when it departs from the axis of the mirror. Also, its scope is quite limited. This is a drawback of the hypertelescope.

The existing interferometers (including the VLTI), use independent telescopes, by mixing their light. But these telescopes are not the same distance of the observed star. It follows that their mirrors are not a same spherical surface. Then, to regularize the situation of a plane wave, must be placed in the path of the light beams, before they do interfere, delay lines, which allow all bundles to arrive at the same time. These delay lines are complex, costly, and limit the number of possible openings (4 for the VLTI, 8 with TAS).

Photo A. Labeyrie

A piece of crumpled and smoothed aluminum foil allows, placed in the Sun, to give an image similar to that of a constellation. Another piece of aluminum foil, placed before the lens of a camera, lets light through a set of small holes (pinhole). We note that:

  • with 15 small holes, we see a halo of light, and distinguished nothing else;
  • with 50 holes, we begin to distinguish the brightest points;
  • with 235 holes, it begins to distinguish a pretty close picture of reality;
  • with 600 holes, we have a very good image, which looks quite like the image obtained without screen pierced (in full aperture).

In this simple experiment, it has not densified pupil, what is the cause of the halo surrounding the image with 600 holes. But clearly, experimentally, that the image improves if it increases the number of fragments of the diluted mirror.

Now, a question arises: what should be the diameter of the components mirrors, to get the best possible quality? Another experiment has led the result: it is much better to increase the number of fragments, rather than their diameter. Equal collecting area, the largest number of mirrors is best:

photo A. Labeyrie

The comparison focuses on two mirrors of same total surface area, the first with 6 mirrors, and the second 600. Since the total collecting area is the same, the radius of the components is R2 = S / 6 π and r2 = S / 600π. Therefore R = 10 r. For example, 1 m 6 mirrors or even 600 mirrors 10 cm! It should not be more difficult to make 600 mirrors of 10 cm, well that the number seems huge, only 6 metre…

At the Center, we see both diluted mirrors images, and there easily one that includes 600 small mirrors gives an intelligible image, which is not the case of the other. On the right, we see a part of the image enlarged, which shows the difference. Rotation contributes little, but we'll see its usefulness.

photo A. Labeyrie

On the right, we see the image without rotation, and there is unusable. With rotation, the difference is impressive, one sees a pretty clear picture. Finally, if one subtracts the background, one arrives at a quite correct picture.

The theory indicates that the resolution of a set of n2 star requires a mirror with at least n components mirrors. To properly separate 1,000 stars, so at least 33 mirrors.

Of course, the resolution (separator power) is even better than the diluted mirror diameter is largest. And this mirror shall consist of a large number of components.

We also saw that the secondary peaks are mitigated for the benefit of the central task by densification of the pupil. However the secondary peaks are produced by dilution. So, if we really want to dilute the mirror to lower costs, densify the pupil. The only downside of this densification and the decrease of the field. Indeed, if one grows the image given by each fragment of a factor γ, the field is reduced by this factor γ, while the central peak is intensified by the factor γ2.


It remains a limitation due to the atmosphere. It contains air bubbles that déphasent the light from a star. Thus, it cannot have a constructive interference as expected.

There are two solutions for this, which are now classics: speckle interferometry and adaptive optics. The second requires important and expensive hardware. We can initially use the first, which only requires a very fast camera, to take pictures during the moments of stability of the atmosphere. The speckle interferometry was developed by Antoine Labeyrie.

Speckle produced by a star

Speckle are instant images of the stars. By adding many instant images, one can build a very good image quality. This is the principle of the method named imaging of speckle (see speckle imaging). The Indian astronomer Arun Surya comes to show that the method is applicable to the hypertelescopes.

Comparison with a Schmidt

A Schmidt telescope has a main mirror ball, larger than the blade from closing. Thus, the light beams arriving from remote (angularly) stars and passing through the blade, arrive on different portions of the primary mirror. For a given star, only a portion is used. But different parts different star imager, and offer a large field.

The angular distance of two stars in a Schmidt can be compared to the two successive positions of a same star, at different times.

Let us not forget that the mirror of the hypertelescope has no overall mount, and cannot be moved. Also, is a star tracking problematic. But if we accept to put mirrors on a surface that is much larger than that will be used and then during the night, the image will move mirrors Mirrors, to be always visible. At any time, only a portion of the fragments will be used.

This provision could be considered to be a waste. Which is true, since a part only is used. Now consider a unused portion at a time. To make it, it would suffice to retrieve the light that it returns! Therefore, to place a second gondola, bearing a similar to the first perspective.

We can therefore build hypertelescopes spherical mirror, and several pods, and lead several observations simultaneously… What was a waste becomes a big advantage! The multiplication of fragments to track an object for a long time without a horse; It also allows to make several observations in parallel.

The hypertelescopes ground

The first

The first hypertelescope was built with a diluted lens of… 10 cm! Do not be expected to make great discoveries with him, it is self-evident.

It is formed on a 10 cm opening bezel, which is placed a pierced mask of 64 0.8 mm regularly spaced holes. The eyepiece of the telescope is developed to the point at infinity, and each hole gives a picture of 0.1 mm in diameter. These images are spaced by 1 mm. It has a densifier of microlenses formed pupil related γ-10. Images of the holes are therefore output 1 mm in diameter, and are practically joined. We have seen that the amplification of the central peak is γ 2, so here it is 100 times.

The first schema hypertélescope

In the diagram, first notice images directly by the hidden left bezel, then after densification right. This illustrates the effect of densification.

This miniature instrument allowed to validate the concept of densifier of pupil, and therefore of the hypertelescope globally.

The mount

We reported earlier that the hypertelescope had no mount. If that were the case precisely, it could point a star only at the zenith, and for a time very short. In order to make the long term poses, it must the device mounted at the home turn to follow the light of the moon. But doing the focal instrument will come out of the field of some fragments, to enter than others. Thus, the diluted primary mirror must be physically larger than what is used at every moment.

Displacement of the pod scheme A. Labeyrie

The limit would be of the order of 1,000 to 1,200 m in diameter, for any stupid reason: there is no largest natural site to install the mirror at reasonable cost!



Carlina Acanthifolia photo A. Labeyrie

Carlina is the name of the first achievement in real size. A first prototype was made at the Observatory of Haute Provence (OHP). It consisted initially of two mirrors, and its purpose was to develop the mechanical part of the set, in fact the gondola suspended from a dirigible balloon. All the systems allowing this carrycot to remain in the air with an excellent accuracy of positioning was to design. It was directed by a set of cables, anchored in the ground, firing the ball in three directions at 120 °. The gondola, for its part, suspended ball, must be able to move slightly to ensure the monitoring and compliance. The system was inspired by what was done at Arecibo in radiosatronomie.

The prototype

Prototype of the OHP; balloon and gondola photo Antoine Labeyrie
All seen from underneath photo Antoine Labeyrie

Reflet544Reflection in one of the mirrors picture Antoine Labeyrie

A third mirror has been added then, making the system more realistic.


Since then, another prototype, full-scale, is being installed in the vallon de La Moutiere, in the Ubaye Valley. The project begins with an opening of 57 m. The carrycot is equipped with a Mertz corrector with two mirrors (used to correct spherical aberration by a mirror not parabolisé).

The site – located at 2,000 m altitude in the South, the Valley of La Moutiere Alps near Barcelonette, responds to these criteria. It boasts a sky without flares with a good transparency and a low turbulence. Its middle part, oriented East-West, close to a cylindrical shape, is sheltered from the prevailing winds and suffers almost no thermal breezes, very favourable to the installation of the suspended structure with focal optics.

Road map of the site (extract from the Michelin map)

The site is located partly in the Mercantour Park, and thus no permanent installation is possible. It is indeed a demonstrator, and not a tool for exploitation. No building can therefore be constructed, and nothing should interfere with wildlife. However, this instrument will remain in place, and will be expanded, until the final draft is implemented, which will take a few years. After removal, there will remain no trace.

Sitemap – the mirror consists of small mirrors to 15 cm in diameter, installed on tripod supports anchored in the soil. The diluted mirror is 200 m in diameter (57 initially), and small mirrors are placed all in a same sphere. The optical pod weighs only a few kilograms, and can be supported by a simple tensioned from one side to the other. The Valley is oriented East-West, the cable is North-South.

The cable – the cable is Kevlar, which gives it great strength and good rigidity. It is 800 m long.

Implantation (schema Labeyrie on background Google Earth)

The nacelle – the gondola, suspended from the cable to a hundred meters from the ground, is guided by 6 small cables, powered winches. Winches motors and stepper. To synchronize the actions of winches, a wifi network has been installed on the site! It allows a computer to control order measures (and possibly to the chamois to connect on…). Naturally, it takes solar power to give the power to this equipment.

Motor West photo Antoine Labeyrie

The nacelle returns light to the South point of the site, where a small telescope in Equatorial Mount to retrieve. The polar axis is oriented towards the north celestial course, but it must also pass through the centre of curvature of the mirror. To do this, simply place the telescope on the polar line that passes by the Center, where it meets the floor.

The performance – thanks to its many small spaced mirrors, this prototype will take the same separator power than the VLTI (small number of large mirrors). A cost had nothing to see… In addition, Carlina gives a directly viewable image, while a classic Interferometer requires a reconstruction of this image by an algorithm after data entry. Get all spatial frequencies, to use the largest possible number of pairs of telescopes. These observations can be done successively, and does not therefore apply to stationary objects, excluding those in rapid evolution.

The resolution of the hypertelescope in its first version, 57 m in diameter, will reach 2 milliseconds of arc: compare with 40 milliseconds of the HST, or 20 times better! In the extended version, 200 m in diameter, will be 0.5 millisecond. Is 80 times better than the HST, and 120,000 times better than the observations of Tycho Brahe…

There where Tycho Brahe saw that a single pixel, we will see 120,000 120,000 × = 14 million! Is a tessellation of 11.039 screens 17 '' side by side, in a matrix of 83 in width by 133 in height (definition of 1,440 × 900 pixels each).

And this is just the beginning…


ELHyT means Extremely wide HyperTelescope. This is the final draft, which is expected to reach 1,000 to 1,200 m in diameter. To do this, will need to install it in one of the few possible valleys on Earth. The most promising site is in India. It should include in the term a thousand mirrors.

Despite these impressive dimensions, the cost of such an instrument would be very reasonable, since all mirrors are open. No mount is to build, and this represents a major savings for the project. The critical point that remains to be validated is the use of adaptive optics, using a laser guide star. This improvement will achieve a resolution much higher than that of existing interferometers. It must also give improved sensitivity and magnitude limit of the instrument, than the current.

An important aspect is the ability to add mirrors to measure. Thus, less greedy scientific objectives can be acquired while the instrument is not completed.

Returning to Tycho: definition of 0.1 millisecond of arc. Is 600,000 times more than Tycho! 278,000 screens 17 ''.

And that is not all…

Space hypertelescopes

To build a space hypertelescope, there are two possible approaches:

  • a rigid set of mirrors, maintained by a large metal structure;
  • a set of small independent satellites, named free-flyers.

The first was envisioned by NASA, which has done studies in this direction (Terrestrial Planet Finder, a beam of 100 m in length with several mirrors). The second is European, and it is definitively adopted.

The Darwin project is a space Interferometer comprising 6 2 m mirrors on a surface of 100 m to 1 km radius. This is not a hypertelescope, since it involves a small number of mirrors.


As free-flyers solution is adopted, it is necessary to solve the problem of stability of the configuration, which was in the other case by the rigidity of the beam. The stabilization in space may be considered by three methods:

  • by micro-fusees (duration of life limited by the amount of fuel);
  • by solar sail (potentially infinite lifespan,) but long reaction time;
  • Laser (a mirror laser radiation pressure can redirect it).

The micro-fusees control has been tested by experience experience Prisma. It has shown the feasibility with the required accuracy.

The solution of solar sail is perhaps not the best because its reaction time to reconfigure the miropir is too long. But it has a considerable advantage: a mirror can be lost, that is, for some reason, its orientation puts it beyond the reach of land-based communications, and it is delivered to himself. Worse, it may interfere with the whole. Or solar sails can be made (in the form of dishes), so that if the mirror is disoriented, the pressure of solar radiation will give in the right direction (approximate) after a certain time.

Piloting by laser is similar to that of solar sail, because it also includes a passive sailing. But instead of solar radiation, this sail is designed to receive a laser beam produced by the central laboratory (which contains the optics). Various studies have been conducted that indicate that accuracy would be better than that of the micro-fusees, for a lifetime not limited.


Epicurus is a spatial hypertelescope project, presented by Antoine Labeyrie, and including 6, then 18, and finally 36 small mirrors of 30 cm diameter, several hundreds or thousands of metres (i.e. a separator power of the millisecond of arc).


Luciola is the successor of Epicurus. It's still a space project of Antoine Labeyrie, about a kilometre hypertelescope. It includes a pupil densifier. Submitted to ESA in June 2007. It is in a phase of simplification to make it operational.

The Luciola drawing Antoine Labeyrie project

The brightness will not be a problem, despite the diluted opening (large holes). Indeed, with 100 mirrors of 25 cm, one reaches the magnitude limit of the HST. Each mirror has a collecting area s = π r2, so the 100 mirrors total S = 100 π r2. This is equivalent to a single mirror of RADIUS R, so surface S = π R2. Tying the two expressions of the surface S: π R2 = 100 π r2. It simplifies by π, and it remains: R2 100 r2, i.e. R = 10 r. 100 mirrors are equivalent to a single mirror diameter 10 times, so 2.50 m more. But the HST is 2.40 m. Luciola will therefore have the same sensitivity, with separator power 400 times larger (diameter 1,000 ratio m / 2.40 m).

Outside the atmosphere, any necessary adaptive optics, and not loss of wavelength. It will be possible to observe from the ultraviolet 120 nm (Lyman α line) until infrared 20 µm.

Methods to increase the dynamics of the image, the coronography, spectroscopy, are usable. It is therefore an absolutely complete instrument that can be performed as well. It fulfilled the conditions to meet Cosmo Vision of the ESA program, which plans the missions from 2015 to 2025. This planning is made indispensable by the duration of the developments (more than 20 years for Huygens).

This concept, as its soil, is inherently extensible. The number of mirrors can increase with successive missions, and optical benches (equivalent Carlina nacelle), may be multiple, brought successively. The number of mirrors could reach 1,000, greatly increasing the sensitivity. This number, with mirrors of 25 cm, reached the same area as one of the telescopes of the VLT.

For imaging of the stars, the first elements will be sufficient. But for exoplanets, it will take a little longer. Starting with the infrared, where the contrast is better and where observations will be possible fairly easily. But for the visible, it will wait until the instrument is more complete. At this time, the combination of Spectra in the infrared and visible will detect life on an extrasolar planet.

Luciola is the equivalent of the Ubaye hypertelescope: a prototype that should already provide extremely valuable information. But this is only a step, a demonstrator. The suite is already possible:

  • IEE: Exo-Earth Imager, of one hundred km in diameter, with a separator power of 1.2 µseconde, either 60,000,000 times more than Tycho;
  • NSI: Neutron Star Imager, reaching 100,000 km, separator power of 1.2 nanosecond, or times more than Tycho…

The way (Finally, to browse again) is unimaginable.

There is an argument to minimize the size of the mirrors. Indeed, if d is the diameter of the mirrors, the mass of the microsatellite is d3, d2 surface. Then, miniaturization decreases by a factor of 1/d acceleration needed to bring the microsatellite in motion (to reconfigure the telescope, the thrust on the satellite is proportional to its surface, while inertia is proportional to its mass, so its volume). Ultimately, the time required to reconfigure is proportional to d1/2.

Use of the hypertelescopes

With a hypertelescope ground, we can better see the stars, their planets, search a life (possibly advanced), galaxies, the distant universe.

Observe a transit of a planet before its star would be feasible, to nearby stars. Spectroscopy during the passage, you can analyze the planet's atmosphere.

With a coronagraph, already began to be (larger) exoplanets imaged. This method would be used with some hypertelescopes. The planet, Earth, is type 1 to 10 billion times less luminous than its star.

All that is ageing today would be course targets: black holes, the active nuclei of galaxies, quasars and micro-quasars (see if the model designed from its properties is much realistic), gravitational lenses, (optical counterpart) gamma-ray bursts…

With a space hypertelescope of 100 km in diameter, made up of 100 mirrors from 3 m, can imager a planet like Earth in 30 minutes of installation. The image helps to distinguish the continents (if there are any), and thus much to his subject. In particular, any change in color depending on the season, so a possible life.

hypertelescope-terre10alView of the Earth

The image above shows how we would see a planet resembling Earth in this instrument of 100 km, if it was placed at 10 AL of us. There are the seas, continents, colors. If there is a change in color with the season, it would be visible. This would not be proof of life on this planet, but it is not far away.

It is obvious that the Lagrange point L2 is an ideal place to install such an instrument. It has a very uniform microgravity, which means very low tidal forces. Yet it is they who are likely to separate the microsatellite.

nanosatelliteA nanosatellite

Supergiants are well resolved from 18 m.

ELT screw Hypertelescope

An ELT is much greater than for a hypertelescope. But of course, its diameter being much lower, its resolving power is also.

A hypertelescope can understand several pods, and constitute in fact several simultaneous telescopes.

Group of stars in an ELT

(Extremely Large Telescope)

The same in a Hypertelescope

Antoine Labeyrie plans until a hypertelescope formed mirrors spaced on a diameter of 100.000 km! Such an instrument, necessarily spatial, would allow imaging the surface of a star neutron (20 km in diameter).



OVLA (Optical Very Large Array) is a Michelson type Optical Interferometer project, proposed by Antoine Labeyrie. Its original version has 27 1.50 m diameter telescopes placed along an ellipse of several hundred metres of a half minor axis. Indeed, the goal is to build a giant parabolic mirror, which remains virtual. The imagining pointing to the sky, light it intercepts projeterait to the ground an ellipse. It is therefore on this ellipse that placed the actual mirrors. The remaining in the centre hole is thus the equivalent of the central obstruction by the secondary, in a conventional telescope. It is obvious that may be considered to partially fill this hole with other mirrors, which would be placed on internal ellipses. This will be an extension of the project.

The instrument is an interferometer, need to recombine the light coming from each of the telescopes in a laboratory, which will be located in one of the foci of the ellipse. It is through elbow homes telescopes that light will succeed.

Telescopes must obviously be every moment good ellipse. However it becomes distorted of course with the diurnal motion. Therefore, that telescopes can move! Each will be placed on a stand with 6 legs, and work (you read) at a speed of a few centimetres per second for an instrumental diameter of 500 m for example. The telescopes should be compact and light. The precision of movement is of the order of the wavelength of light, but interferometry lay a little this constraint.

The 27 = 33 number allows grouping by 3 telescopes for consistency; then each group by 3 again; and finally these clusters together.

The system will use a pupil densified, enabling it to achieve a resolving power of 10-3 to 10-4 seconds of arc. In addition, it directly Gets a usable image, it is unnecessary to rebuild by calculation.

The originality of this configuration lies in its flexibility: depending on the type of object that you want to observe, can quickly configure the appropriate ellipse. This distinguishes it from a system of the VLTI genre, in which, because of the delay lines, the only possible locations are determined at construction.

The telescopes are Gregory. The mirrors will be 1.52 m in diameter, thin ordinary glass, supported by 29 actuators. They will open at F/1, 7. These are meniscus, to limit the weight. Secondary 75 mm in diameter, and a tertiary mirror returns the beam towards the elbow home.

The mount is a ball, which has many qualities, including that serve also shelter to mirrors. These telescopes have not dome! What makes a huge saving on the construction. The accurate pointing is ensured by three rollers that support the ball and turn at the same time. It is the lightest telescope of the world, because it weighs less than one tonne for a diameter of 1.52 m.

This instrument will have a separator power from 100 to 1,000 times better than existing telescopes, HST or VLT (with adaptive optics). Small bodies, such as comets become imageable remotely. But the surface of some stars (Cepheids, RR Lyrae) is also visible. Further, the active nuclei of galaxies. But the most spectacular probably concerns the exoplanets, which will be visible by coronography or nulling interferometry.

Source: Seminar Hypertelescope