Last night in Paranal

Paranal landscape at sunset (photo Roland)

As almost each evening, I am contemplating  the magical transformation of the Paranal landscape into a subtle collection of reds, oranges  and browns while the sky turns to deep blue. This is also the time of question and expectation: what seeing, how much ground layer and wind tonight ? The program is ready, we will start with the quasar and then move to other interesting objects before finishing to the now obligatory on/off sequence, now nickname snake.

But as a sailing boat does not decide on the direction and strength of the wind, observation at a ground based telescope has to cope with some part of unpredictability, some hasard, which is inherent to the atmospheric behaviour.

Indeed, at start, we were not able to point to the quasar because of pointing restriction due to strong wind from the East. This was unexpected  because most of the time the strong winds come from the North and thus two days ago we changed on purpose to a South  target to be on the safe side. We had to wait that the object pass meridian and in the meantime perform a snake sequence to measure the AO performance. Seeing was terrible, almost 2-3 arc second but the system manage to put it back to 1.2-1.5 arc second, not great though in absolute terms.

When it was possible we then moved to the quasar, but the measured image quality was even poorer. Even if the AOF was still closing all the loops, the image quality  indicator went off the high limits and disappear from the monitoring screen ! We still performed the expected exposures but with such a poor image quality we were not sure what will be coming out of it.

Outside, wind was blowing like a hell and the ground layer vanished, AOF was desesperatly  trying to keep the tip/tilt star in its box. In an heroic attempt, we keep trying to perform some measurements all nights, but at 4 a.m we decide to leave the boat.

The Ultra Luminous Infrared galaxy ESO 148-G002 observed with and without AO (images Peter).

In the meantime, as a consolation, Peter processed the date obtained the previous night on the ULIRG and produced a beautiful image that we are happy to share with you. Look to the finer detail and improved sharpness when comparing with an early observations obtained a few years ago at the science verification.


After a good sleep and a quick breakfast, it is time for all of us to debrief at the meeting room of the residencia. Despite the poor weather of the last nights, we have been able to perform all key tests and improved the system. A first evaluation of the performance demonstrate that our expectation are met.  What was almost a dream 16 years ago is now a palpable reality.

With our data in hands we can say that image sharpness that were extremely rarely measured in MUSE or even impossible to obtain, will be obtain in normal conditions, say more than 50% of the time. Moreover, in average, the worst the atmospheric conditions, the better the improvement, resulting in a much higher probability for the science user to obtain excellent image quality, even for very long integration. Of course, this must be confirmed in the long term  by using the system in various atmospheric conditions, but it is likely that the combination of MUSE and AOF will improve the quality of the produced MUSE datacubes and, ultimately, boost the scientific productivity.

One of the most complex systems ever build for an optical ground based telescope is now at our disposal. We are all excited  and impatient to use it to question the Universe and get more hints and insights to try to answer its many mysteries.

On behalf of present and future users that will have a chance to get access to this great facility, I would like to thank all of you, and you are many,  that have been part of this great technical and scientific development. It is a unique and major collective achievement, you can be proud of it.

This last night marks the end of the MUSE plus GALACSI commissioning. It has been a privilege and an immense pleasure to share with you, commissioning companions, these two runs. There are no successful scientific and technical achievement which is not also a human successful aventure, and this one is not an exception.

Note also that, August 2, ESO will communicate to the world a press release with some of the great images we have obtained.

I will close this blog for some time. It was great to share with all of you this moments. This is already the second episode of the series: first light and commissioning of MUSE stand alone was the first one (see here). But wait, this is not the end, in hardly a year we will be back to continue our adventure, with the expectation to push forward the system to another level: the so-called MUSE narrow field mode will be coupled with a full laser tomography adaptive optics system. The aim is to improve the sharpness of MUSE images by almost an order of magnitude in a small field of view and we expect exciting results when scrutinising astronomical objects.

Stay tuned, we will be back !

Time for me to say goodbye. I am just a poor lonesome astronomer … (photo Simon).

The big catch

Clouds over the pacific at sunset (video Johann).

Tonight, I have seen it. Not only me, most of us have seen it. Like you, I had doubt that it exists. There were rumours but nothing more. Some time ago I met an old man who told me that he has seen it. He gave me a lot of details, but can we trust a foreigner ?

Peter and Gerard are looking to the big snake (photo Roland).
Johann is wondering if it is dangerous (photo Roland).

The night was clear. After some observations of a quasar fields, we were forced to leave the place because of risk of collision. We then move to south and catch NGC 5694, one of the usual southern snake species. At 6:25 UT the wind went down and Peter suggest that we move to a ULIRG, an ultra luminous infrared galaxy. We found a nice one, with a lot of jets.

Benoit is already hypnotised by the monster (photo Roland).

We were busy measuring the complex gas structure while I notice something moving fast in the background. We slowly approach the place and suddenly, I saw it. It was slithering in front of me, huge, almost two arc seconds at the top and less than 0.7 arc second at bottom.  On its off state, it was almost filling all our field of vision while on its on state it shrinked in a minuscule spot. But the wind got up again and it started to become dangerous. We just had time to took a picture before leaving the place.

At the end of the night we safely went back to the residencia and went to sleep, trying to forget the monster.

The laser show must go on

One night at UT4 (video Johann).
Thierry and Lutz are studying the freshly obtained data sets to develop image quality indicators.
Study atmosphere in the control room.

Tonight we perform many tests to check the functionalities and the robustness of the system. We also did as usual a  number of on/off sequence to measure the performance of the system under different atmospheric conditions.

At the end of night we had performed most of the planned tests and accumulated tons of data cube that have filled all the available computer disk space.  And, icing on the cake, the problem with the offset sequence was solved thanks to Joel, Johann and Pierre-Yves who spend their last 5 a.m neurons to demonstrate that a cos(delta) was missing in the sequence.


The revenge of the snake

Shooting in clear sky patches (photo Roland).
The AO on-off sequence, aka the snake, is back.

After the first chaotic night, the team is seeking for a revenge. At the start of night we launch the lasers, ignoring Paranal’s desperate attempt  to intimidate us with some clouds.


Captain Pierre-Yves is at the command of the deformable secondary mirror (DSM). Right hand up: “too much tilt on the right, please turn left” (Photo Roland).

A few clicks and the familiar images of the four laser pineapple slices show up, the tip/tilt star is catched within a second in the field selector box, loops are closed and after a few seconds of suspens … the snake is back !


Then we follow our long program, did a lot of tests, all successful, and get many great data sets.

Simon, our snake hunter, has just flushed out a new specie (photo Roland).

Even in these less than ideal conditions, we can measure the improved performance given by the system. We also obtain a beautiful on/off exposure of a planetary nebulae. Great images, but sorry, you will have to wait for the ESO press-release to see them (expected late July, early August).

Harald and Joel discussing cosmic laser collisions (photo Roland).

All the team is focused to maintain  the machine at full speed. Gigabytes of data are coming out. Outside the laser ballet illuminate the sky.

It is 6 am at the clock, the pale glow in the east announce the sun rise. Already ? What a great night !!

Happiness  in the control room for this smooth and successful night. In the foreground: Fernando, Yara and Harald (photo Roland).

Second commissioning run starts: back to the laser game

The direction is clear but the road can be bumpy (Photo Roland).

After a short break of 3 weeks, we are back in Paranal for the second and last commissioning run of MUSE with the Adaptive Optics Facility Ground Layer Mode.

We are a large team of 14 people (from ESO Garching and ESO Paranal in addition to a few MUSE consortium members), but still not many with respect to all those who have been worked on the project for so long.

While the sky turn from purple blue to red as we approach sunset, the team, full of energy, is moving to the VLTI control building. Nous voilà … lasers, wavefront sensors, MUSE be ready, sky is ours !

However, sometimes, you know, the reality seems not to obey our will. Time to time, indeed, there seems to be a conspiracy to prevent everything to go as planned. We all have experienced it. It usually starts with a small sign, a sort of tremor of destiny, for example the alarm clock did not ring because you forget to change the battery. Then it is followed by other imperceptible reality creaks: the bread gets burned in the toaster. From this stage, everything will go wrong the rest of the day: you can be sure that if you have to leave to the airport, the time to find the keys of your car which have been left in the trouser pocket which is currently in the washing machine, it will be already too late to have any chance to catch your plane…

This is mostly what happened tonight. A small extract of tonight’s report prepared by Pierre-Yves is a good summary of the situation:

  • 01:00 UT: start of the MUSE GALACSI commissioning
  • 01:20 UT: MUSE-GALACSI acquisition on a star field – the acquisition went well: no issue – but it seems that the OB was not the right one… to be started again
  • 01:40 UT: new acquisition (NGC6366) – the acquisition went well: no issue – the goal is to have 6 successive open-closed loops to measure performance, with a last one using the commissioning camera; after the second cycle, SPARTA went into a strong error (never ever experienced before) requesting to reboot SPARTA – due to this error (more likely), the DSM was rejecting all commands; Claudia went to zenith to cycle the DSM but got a hard limit switch error on the rotator – Paranal mechanical and electrical support called to get out of this situation – it seems that a hard limit had been set, limiting the rotation angle of the rotator to +/-43 degrees which does not allow to go to zenith when in pupil tracking mode – to be further investigated
  • 02:00 UT: in the meantime, Jose Luis went to the telescope to check the cooling issue – he noticed some coolant flow warning not only on LGSU1 but also on LGSU2 – he increased the flow on the secondary circuit of the Heat Exchanger – he will refill the tank of the secondary circuit tomorrow (missing some coolant) – we check how Laser Units are behaving
  • 04:05 UT: start again with MUSE GALACSI tests… – new acquisition (M75) – the acquisition went well: no issue – the goal is to have 6 successive open-close loop to measure performance, with a last one using the commissioning camera; after the second cycle, the main process of SPARTA died; HO loop was closed and everything else OK; Gerard suspects this might come from the FITS Header data collector; SPARTA main restarted; data collector of SPARTA parameters disabled.

Not many sky data tonight: no planets, no galaxies, little stars and zero planetary nebulae. We still manage to obtain a few on/off sequence to measure the improvement of image quality. Still great, ouf !

“The wise man learn from his mistakes” wrote Confucius. At the end of this long night, we feel much wiser than yesterday. Times to go to bed.

Blinking flash in the dark

Paranal sky has been at its best for our last night of this run: perfectly clear and with enough ground layer to enjoy the power of the AOF. We focused on  performance assessment, looking to star’s cluster at various places and measuring the improved image quality when we turn on and off the adaptive optics system.

We perform many on/off sequence of observations to measure the improvement of image quality brought by the AOF (photo Roland)
Core of the galaxy cluster Abell 2204, showing structured gas emission around the Brightest Cluster Galaxy (near the bottom of the image). A small horizontal arc is also visible above and is due to gravitational lensing (image Johan R)

We also took some longer exposure to measure the laser background impact.


A first quick analysis of the system performances gives us confidence that we will achieve a significant improvement of image quality most of the time. This will open a new discovery space for MUSE.

Thanks to AOF and Paranal experts, the robustness of the system has been much improved during this first commissioning. There is nevertheless still a number of open actions and we will be back in two weeks for another commissioning run.

The AOF magician Johann is invoking the laser power (photo Roland).

It is time for us to leave. Thanks to all of you – Elena, Fernando, Gérard,  Joel, Johan, Johann,  Pierre, Peter,  Ralf, Sebastian – for all the work performed in a very nice and constructive atmosphere. The commissioning team is tiny part of a group of many talented people which have been working hard for many years to achieve this fantastic results, the list is too long to be written here, but I want to warmly thank all of you. Thanks also to the Paranal staff for its hospitality and support.

It has been a great experience to share and we enjoyed so much the laser show in the magnificent sky of Paranal.

Laser traffic upon UT2 (photo Peter).

Stay tuned, we will be back on July 13th.




A perfect night


Sunset at Paranal, the start of a perfect night (photo Roland)

At sunset, Paranal sky shows its intention to cooperate, and indeed, it did all night. After a look at the beautiful sunset, we did not even wait for the sky to be fully dark and could start the big laser show.

The laser is of course the most obvious part you see when you are on the platform. But it is only a subset of the AOF, a fantastic master piece of technology.

A typical observation always start by pointing the telescope to the target and then shaping the primary mirror to the correct optical surface to produce sharp images (the so-called active optics system). Then the laser show starts. First the beam is propagated and hits the atmospheric sodium layer at 90 km above, producing four artificial stars. The four wave-front sensor cameras are then pointed to the laser guide stars. The field selector is then moved to the selected natural tip/tilt guide star and locked to it. The full analysis of the atmospheric turbulence is then performed using these 5 guide stars. Information is processed in real time, 500 times by second, and the computed correction is sent to the secondary deformable mirror. This mirror of 1 m diameter and only 2 mm thickness, is levitating, linked to its support only by active magnets. The deformation of the mirror then corrects the light entering MUSE, producing even sharper images.

The closing of the numerous loops to control this complex system is performed in 2 min. Impressive. And then you can monitor the system on the numerous screens.

AOF GALACSI monitoring screen displays in real time the shape of the deformable secondary mirror (lower right), the tip/tilt star (lower left) and the four laser wave front sensors (upper-left panels). Video Roland.

After performance measurements on a globular cluster, we went to some deep sky exposures to evaluate the impact of the lasers on the background. There is no moon, time to go to the platform and contemplate the millions of stars of the Paranal sky. These four bright light sabers pointing to the target add a surrealistic touch to this fascinating landscape.

Laser shot pointing to M68 seen from the UT4 dome (photo Roland)
Reconstructed color image of a short exposure (1 sec) of Saturn (image Peter).

This night was very productive, we obtained many gigabytes of excellent data over various targets. In the middle of the night, we even did a stop to greet the lord of the rings.

At six in the morning, after this successful long night, it is time to take some rest before the next one.

Cloudy outside, bubbles inside


Cloudy night at Paranal (photo Peter)

Many clouds and gray sky at the start of the night did not look very promising. We nevertheless launch the laser and close the loop on some bright tit/tilt star to perform further tests. We also continue to investigate software issues.

Celebrating first light (photo Diego)
Chair crisis in the VLTI control room (photo Roland)

But Paranal sky did not want to cooperate and remained cloudy all night. Although the show of the lasers shining through the various layers of clouds was nice, we took this opportunity of free time to pop a few champagne
corks to celebrate the first adapted light of MUSE. Suddenly the grumpy atmosphere became much more cheerful. Many

The question is “why is the field selector at this location” (photo Roland)

thanks to the Paranal chef and his staff for the delicious petit fours and mini cakes.

The rest of the night was spent on functional tests, software debugging and long discussions  about the GALACSI field selector, the elongation of stars and various others items.

Deep thinking about center of rotations, offsets and directions. Joel knows the solution (photo Roland)

Let’s hope for a clear night tomorrow.

How is your ground layer today ?

“Did you get good seeing last night ? Excellent at the start,  but at the end very poor ? I wish you clear sky and good seeing for this night …”

Nice sunset, no wind, time to go ! (photo Roland)

This is the typical discussion you can hear at any ground based observatory site because seeing is the obsession of all observers. However if you want to be trendy, you have to ask about ground layer. Nowadays  if you happen to meet some MUSE observers you may hear something like this: “did you say 80% of ground layer ? that’s great, let’s go to the next target then”.

The great laser show over the milky way (Photo Roland)

The ground layer is the part of turbulence layer (the first 500 m) which is beautifully corrected by the AOF GALACSI Ground-Layer mode, just before entering MUSE. And if it happens that most of the turbulence is in this layer, then the improvement of the image quality shall be good. That was indeed the idea 15 years ago when discussing the AOF project and its coupling with MUSE.

Photo Johan R

This second night looks favourable. The wind stopped and a few clouds stay a bit at sunset, just the time to take some nice pictures. Soon after dark time, the laser show start. Tonight’s show was quite spectacular, it is an unforgettable experience to stand over the platform looking at the milky way spiked by four blazing swords.

Back to the control room, while some are performing additional tests to find solutions of the few issues identified at the first night, others are producing the first analysis of the performance tests, on the globular clusters NGC5694 and M75.  In both cases we can see the two direct gains of the AO: the many stars in the cluster get sharper images, and some of the faintest stars in the fields reveal themselves much more easily.

A stellar field in the vicinity of the globular cluster NGC5694, with and without the adaptive optics correction (Image Roland)








The center of the globular cluster M75 in open and closed loop (Image Johan R).
Planetary nebula IC4406 (in Lupus) 2000 light years away (Image Johan R)

Overall, it was a very good night. We obtained data of excellent quality and learned a lot about the system.

Welcome back

Joel and Johann K are inspecting the GALACSI AOF module in the integration hall at ESO Garching (photo Roland)

Three years after the end of MUSE commissioning, we are back to Paranal for another exciting adventure, the coupling of the Adaptive Optics Facility with MUSE.

But what happened during these three years ?  After its very successful commissioning, MUSE was immediately offered to the community and it became so popular that it is currently one of the most demanded instrument of all VLT instrument suite. It has since produced many discoveries, published in scientific journals and some of them have been reported in ESO Press Releases.

Cloudy sunset, high winds at the start of the night (photo Roland)

In parallel, ESO engineers have been working hard to transform UT4 in an adaptive optics telescope. The so called Adaptive Optics Facility (AOF) and its GALACSI module, will correct for the blur produced by the atmosphere, producing sharp images that will boost MUSE performances.

Kick-off meeting at the Residencia to prepare the night program (photo Roland).

The coupling of these two complex systems, AOF and MUSE, will produce one of the most advanced and powerful technological system ever built for ground based Astronomy. Improved spatial resolution, i.e sharper images, will not only allow MUSE to scrutinize in even more detail objects in the sky but will also improve its detection limits to fainter objects.

An experienced team of MUSE observers has joined the talented AOF team and the crew at Paranal for this first commissioning run of 5 nights.  AOF has been independently tested by the AOF team over the past months. Everything is in place, but the commissioning of such a complex system is not straightforward, nobody knows how long it will take to get it ready for science. Expectations are high.

We finally meet the four lasers … very impressive ! (photo Roland)

For our first night with the new system, we planned a battery of tests for the performances “end-to-end”. The idea was to measure the gain offered by the AOF on MUSE for given atmospheric conditions. Despite the strong winds and some small technical issues with the telescope very early in the night, we were soon rewarded by the spectacular images obtained thanks to the

The slow guiding system oscillations show the change of image quality when we switch the AOF on or off (photo Roland)

AO!  Playing with the switch “On / Off” we could clearly measure the gain in image quality.

While the very long winter night is approaching the end, we face a number of technical problems. Debugging such a complex system at 5 am is not the easiest task.

Despite these small mishaps, this first night has already given some interesting results that we have to digest. Robustness has still to be improved but the performances seem very good.

Stay tuned, for some nice images tomorrow.