LMAO - First Light

LMAO - The LEGO Miniature Astronomical Observatory

One of my most recent projects has been the construction of a miniature remote observatory for the roof of my parents house in Melbourne. After spending hours trying to design how this this would work, I decided that I didn't need to design the thing from scratch. Rather I could use one of the best prototyping media out there - LEGO! Due our family's vast reservoir of LEGO, this made life a lot easier.

The Telescope

As far as size goes, this will probably be one of the smallest remote observatories around. The mirror diameter is 76mm and it has a field of view of around 1 square degree. It was originally a low-cost show piece mass produced by Celestron for the International Year of Astronomy 2009. Retail value - €50. While the telescope itself has reasonably good image quality - being diffraction limited at 2 arcsec, for the sake of slewing speed, I could only build in a tracking accuracy of 3 arcsec. Hence the finally images will be a little blurry at the pixel level.

The LMAO 3 inch telescope in its LEGO Hooker mount.

The telescope itself is a replica of the original mirror telescope designed by Isaac Newton and the mount system is a variation of the Hooker mount as used by Edwin Hubble in the 1920s.

Once the telescope is up and running, I'll start work on documenting how all the components were put together so that others may also construct their own LMAOs and we can start a network of miniature astronomical telescopes accessible to the public and placed around the world.

First Light

Obviously there is still a lot to do before the LMAO is finshed and free available for public use. However, as proof that this little observatory will actually be able to see something, I have include the first light image below. This is a 10 second exposure at ISO 400 of a random star field which I used to check the focus. As you can see, quite a few more stars are visible than with the naked eye. The goal is to be able to image the many nebulae that the southern hemisphere has to offer.

Iimage of a random star field, used for checking the focus of LMAO - 10 seconds at ISO 400

Supernovae

Supernovae

On the night of the 25th of February 2014, the Leopold Figl Observatory with it 1.5m telescope came back on line after its winter hibernation period. I was lucky enough to be the first to use it. Thankfully everything work perfectly and Prof. Zeilinger and I were able to take many images of two (at the time) famous supernovae - M1, The Crab Nebula and SN2014J in M82.

M1 - The Crab Nebula

The Crab Nebula is the remnant of SN1054, a supernova that was recorded by the chinese in 1054 A.D. Located only 2 kpc (6500 ly) from Earth, the supernova was visible during the daytime. It has since dimmed and its remnant was only rediscovered in 1731. 

Its dimensions are a little over 4 x 6 arcmins, hence it doesn't quite fit into the FOV of the Leopold-Figl observatory. We observed the nebula in the following four bands: Ha, OIII, Bessel V and Bessel B. The above image is an RGB composite with R=Ha, G=OIII and B=Bessel B.

SN 2014 J in M82

SN 2014 J was discovered by a group of undergraduate students at UCL on the 21rd of January 2014. It is the bright point at the right edge of M82. We observed the Type 1a Supernova 5 weeks later and 3 weeks after peak brightness. The spectrum of the supernova shows that SN 2104 J lies behind a large amount of interstellar gas. Type 1a supernovas are the most common in the universe and are used as so-called "standard candles" to measure distances to galaxies.

Our image is an RGB composite with R=Bessel R, G= Bessel V and B=Bessel B. The dimensions of M82 are around 11 x 4 arcmin, so I mosiaced two separate images together to get the whole of M82 onto a single image.

Repeating Galileo's work

Repeating Galileo's work

Watching the moons of Jupiter

Introduction

Observing the movement of the Jovian moons was one of the first things Galileo did with his telescope in 1609. It helped pave the way for the adoption of the heliocentric model of the solar system. Due to the short time-scales involved and the relative closeness of Jupiter, this is a relatively easy and satisfying one-night project for any amateur astronomer.

Equipment

1x Telescope (Sky-Watcher 76mm mini dobsionian)
1x Modified webcam (MS LifeCam Studio)
1x Laptop (Acer AOD270 netbook)

Software

SharpCap
iPython notebook

Method

Step one was to find out when Jupiter began rising with Stellarium. I then looked out the window in the stair-well in our block of flats and found the brightest star. Pointing my mini-dob at it I confirmed that it was indeed Jupiter. I then removed the eyepiece and put in the webcam. I used SharpCap to take several still frames. This process was repeated every 20 minutes until Jupiter passed out of view. In total I managed to get 8 usable frames.

Figure 1: A comparison of the first and last images of the observing run. Note the shift in alignment with respect to the telescope

The next day I wrote a small script in iPython notebook to find the pixel coordinates of the centre of the moons and crop the image around them. The main problem was that, due to the rotation of the earth, the alignment of the moons moved in each frame. Thankfully the python module scipy has a function ndimage.rotate() which allows the user to rotate an array by any amount and resample it onto a new grid. Thus I found the angles of each alignment and rotated the images accordingly.

The final stage was to add each of the cropped, rotated and centred frames into a GIF image. For this I used an free online GIF animator. The result is seen below.

Results

Figure 2: A GIF showing, from left to right, the movement of: Callisto, Io, (Jupiter), Ganymede. This sequence was taken over a timespan of 208 minutes (3.5 hours)

Unfortunately seeing wasn't very good during the observation run. Hence why the images wobble and the size of Jupiter appears to change from frame to frame. However regardless of the seeing, the change in position of the three moons is clearly visible. Io (second from the left) was at its farthest from Jupiter and therefore was moving more along the line of sight than perpendicular to it. Consequently Io only moved by about 10 pixels. Ganymede on the other hand was approaching its closest point to Jupiter and so moved 25 pixels in the 208 minute observation run.

While not particularly useful for a science, this small 3 hour data-set is useful for understanding how modern astronomy began. I know I can read about the moons of Jupiter in any textbook and see them in Hubble images, but I still get a kick out of seeing the change with my own eye (and webcam) in real-time.

Clear Skies!

The Moon with a 76mm mini-dobsonian

The Moon with a 76mm mini-dobsonian

Just for fun I took some picture of the moon from the stair-well window with my little mini-dobsonian and the modified MS LifeCam that I got for Christmas. The seeing was fairly crappy, being winter and all and living inside a big city. However the HD resolution of the webcam does a nice job at bringing out some of the fainter crater details. I used the wavelet tool in Registax 6 to sharped the image from the 10th of January.

10th of January 2014

12th of January 2014

Getting to know the Nordkuppel

Getting to know the Nordkuppel

On the 3rd of December 2013 I was fortunate enough to have the University of Vienna's 0.8m Nordkuppel telescope all to myself for a night. Before the humidity rose beyond acceptable levels (80%), I managed to take pictures of M33, the Triangulum Galaxy in the R, V and H alpha bands as well as the SII and OIII bands. The first 3 images showed quite a lot of detail, while the later two weren't exposed long enough to reach a usable signal-to-noise ratio.

M33 - The Triangulum Galaxy

The red and green channels are used for the R and V images, while the blue channel shows the H alpha emission. As H alpha traces the excited gas, it can be used as an indicator for recent star formation. The youngest stars are blue, hence why I used the blue channel to show where these star forming regions are located. The dust lanes in the spiral structure is also visible as dark patches in the otherwise yellow background of field stars

M15 - Globular Cluster

In the final 15 minutes of my observation time, I swung the telescope around to look at the M15 globular cluster. This image was taken in V band.

Notes on the RSR of XERO

Notes on the RSR of XERO

During the IAYC 2013 I gave one of my participants, Alex Barr, the arduous task of determining the relative spectral response (RSR) of the CCD that is used in the Xbox Live Vision Camera. The motivation behind this is the comparison between the RGB channels of an Xbox Live Vision webcam and the standard BVR photometric filters. By knowning the spectral response of each on-chip filter in the CCD we will be able to determine the conversion factors between photometric values in the literature and catalogue data gleaned from the XERO Southern Hemisphere Survey.

In order to determine this RSR Alex observed Vega with a 70mm refracting telescope. The Xbox webcam, together with a Star Analyser 100 lines/mm grating, were attached to the telescope in order to record the spectrum of Vega. 

Figure 1: Single spectral image of Vega. Vega can be seen at the far left and the first order diffraction spectrum is the rainbow streak stretching from the centre to the very right.

After reducing, aligning and stacking the raw images we measured the intensity of the spectrum along the length of the image. A single raw frame of the spectrum is shown in Figure 1. The diffraction grating of the Star Analyser 100 splits up the wavelengths nicely. With the IR filter removed the sensitivity of the red channel is vastly increased, however there appears to be an IR leak in the green channel. This will prove to a challenge when the time for photometry comes.

The relative spectral response curve was easy to extract, callibrating it was the difficult part. We chose to use the hydrogen alpha and beta lines as reference points. These are shown in Figure 2. By using these two lines, we found a resolution of 1.5 nm/pixel. The final spectral response curves are shown in Figure 2.

Figure 2: Spectral response for the inbuilt RGB filters in an IR capable Xbox Live Vision webcam

We confirmed the spectral resolution by assuming that the wavelength range of the blue and green channels wouldn't be too different from the wavelength range of the standard Bessel B and V filters. Assuming a FWHM for each of the Blue and Green channels of 95nm and 90nm respectively we found a spectral resolution of 1.4 nm/pixel. Therefore we concluded that we had indeed calculated an accurate spectral resolution by using the hydrogen absorption lines.

Comparing Xbox Webcams with and without the IR FIlter

Comparing Xbox Webcams with and without the IR FIlter

While on a camping trip to Croatia this year, I took two Xbox Live Vision webcams with me. In February 2013, I had installed one on the roof of my parents house in Melbourne, Australia without the infrared filter, so during this trip I wanted to see how much of a difference the IR filter makes to the sensitivity of the webcam. 

I used SharpCap to take 10 second images of several regions of sky with both webcams. I had removed one of the infrared filter with a swiss army knife. 

Raw Images

Figure 1: Raw 10 second exposures of Cygnus an Lyra with and without the IR filter in place

With the IR filter removed, the wavelength range available to which the CCD is sensitive is almost doubled (from 400nm-700nm to 400nm-900nm). Consequently more stars are visible. However, as the sky background emits at all wavelengths, the gain in sensitivity is barely noticeable. Other object which emit more in the infrared regime are also visible - for example the tree in the bottom left corner.

Reduced Images

 Figure 2: The same images after the dark current had been removed.

The sensitivity increase is much more noticeable once the dark current is removed. The amount of stars visible has almost doubled. This step was done using Registax 6 and a simple dark subtraction.

Annotated Images

Figure 3: The constellations Cygnus and Lyra and information about several of the stars

The annotated stars in the left panel are visible in both images, whereas the star named in the right image is only visible once the IR filter has been removed. This is because the star is an M class dwarf whose SED peaks in the IR and drops off sharply in the green and blue bands. Almost all of the stars visible with the IR filter are blue stars, more massive than F class. In contrast, without the IR filter the there is a much more even spread over spectral types. The images taken through an IR filter are limited to around 4th magnitude stars, while the removal of the IR filter, the limit reaches 6th magnitude for late type stars and 5th magnitude for early type stars.

Source Extraction on Images

Figure 4: Comparison of the amount of stars found by the extraction algorithm in the Astrometry.net program

The amount of stars extracted from an image also increase after the IR filter was removed. This both increases the amount of stars for the XERO catalogue and helps the programme Astrometry.net find a better WCS match for the coordinated of the image.

In summary, by removing the IR filter from the Xbox Live Vision webcam, the webcam moves from a curiosity to a usable CCD for taking images of the night sky. Images taken with the webcam are on the same sensitivity level as the human eye, and comparable to the BRITE satellites. The downside is that Red filter photometry is no longer roughly comparable to the Bessel R filter.