Yes, I know (and I am ashamed) so careless that I have this Blog. But I promised myself that that will change, so for the purpose of amending my attitude, I will try to write a while, at least once a week ... Even if I'm not able to fulfill my promise, at least I try ... Let us begin ... On Saturday March 19, I decided to do some tests with this technique has been called speckle interferometry and say that most know, is the solution to cancel the capricious dance of our atmosphere, allowing us to measure peer access closed, which is said closed ... Well, almost as closed as the theoretical resolution limit of the instrument employee permits. So I'll try to explain, what if you gave my first experience "speckleriana" and the conclusions that I arrived after 3 hours of hard work .... This is my chronicle particular comment ... Date: March 19, 2011 Time: 2011.214 Start Time: 20:50 UT End Time: 23:50 UT Temperature: 12.9 º C and 11.9 º C Humidity: Between 86% and 89% Wind: Calm Moon : Fill (perigee) the equipment used in this first test session with the speckle interferometry method consists of a SC LX200R telescope of 254mm aperture f/10 equatorial mounted and is fitted after microfocuser a Vixen Flip Mirror in apart from one eye 20mm grid is inserted an HR16 Atik CCD with a 1392x1040 pixel chip of 6.45 x6, 45 microns, which is attached a Celestron Barlowx2. For these initial tests do not use any type of filter photometric or cutting. The objective of this session is to try shooting the pair to 1352 whose components had magnitudes 9.42 and 11.48 in 1991 an angle of 40 º and separation 1.1 "... The reason for the choice of this pair, are based on the following reasons: - During the observation time, the constellation Uma where is located the pair, is located in the most optimal situation for the observation from the OANL. - It's a couple that, as regards magnitude, begins to be in the top of my opening, using speckle technique means. (The diameter of the target mark to the maximum quantities that can work with exposure times as short as those used in interferometry) Why start with the easy? Let's find out the limits of work as far as magnitudes are concerned, I said ... - The separation seemed affordable to the team, whose theoretical resolution limit is 0.46 "... (That means about three pixels of separation for this pair ....) Also should compensate a little added difficulty of magnitudes. Although at first I would have liked to use the same Florent Losse method used for calculating the calibration constants in the end I did it with a combined ... And, the calculation of the rotation angle of the camera is simple to calculate, taking a few simple traces stellar engine stopped. Another thing is to calculate the pixel scale ... To do this correctly you should spend some more time, doing shots of various parts of the sky, the primary focus first and then work settings and to describe accurately the focal length , something unthinkable for SC (and expressly discouraged by Florent) ... so I decided to use the traditional method using pairs of calibration. Determined a rotation angle of -1.06 º camera based on averaging the angles obtained by Drift Analysis function, from stellar traces some of the stars near the work area. Here you have a sample image. This is the trace of Dubhe (alpha Uma), one of the stars used for determining the rotation angle of the camera.
To determine the pixel scale, have taken pictures of double calibration STF 1603 (AP83.28 th Sep 22.31) and STF 1349 (AP Sep 19.22 165.62 º). And as you know I'm a little curiosón, I thought ... Why not try to make Reduc reduction to calculate the pixel level, first through the traditional method of reduction, after using an autocorrelation (in the purest style speckleriano ") and compare the results obtained using both methods? Well, in that I got, and here are the results ... Traditional Reduction ... ... ... ... 0.1886 "/ pixel reduction Autocorrelation ... 0.1887" / pixel As you can see, the results differ by 1 ten thousandth of a second of arc ... Incredible No? The results (by the traditional method ...) were averaged, obtaining for the optical configuration used, a scale of 0.1886 "/ pixel, which means working with a focal length of 7203mm (f/28.5) ... which is almost triple the original focus of the instrument! I am sure that when you clean a little technical, it is not mission impossible to increase this focus. Perhaps adding a further Barlow current optical system or projection images are acquired eye. For you to see them, here are some pictures interferometric STF 1603, one of the pair used to calibrate the pixel scale ... This individual takes, you can see the symmetry of the diffraction spots on both stars. In 
this one can see the torque correlograms obtained with NR. 
But ... How do we know what high school?. You will ask ... Well simple ... is the one with the highest peak brightness in the image on the right ... These are the things that has the power to perform an autocorrelation Crusade or "Cross-Correlation." Another benefit of further reduction, allowing us to save the reef from the ambiguity of 180 degrees making the secondary in the correlograms. Point out that images can be seen in this post, there are pictures Speckle in the strict sense of the word ... The shots are made with the technique has been called "Lucky Imaging" .... With the diameter of our goal and the "high" integration times used (due to the need to achieve a minimum signal to noise ratio in our images) and are marked by the characteristics the camera, you can not get images of spots clearly defined and individualized. Thus, the images we get are images Speckles to the "gross" or what is, images of "Supermanchas" or Florent Losse calls with elegance "SuperSpeckles." You can read his explanation on the subject in this thread ... " http://www.blogger.com/% 22http: / / www.asociacionhubble.org/portal/index.php/foro/viewtopic.php?t=36153 " Well, as I have a bit boring with the issue of calibration, I will show the final result of tests performed on the pair to 1352, presented in 1991 to remember a PA of 40 º and separation 1.1 ". It should be noted that according to historical measures, this pair has an annual variation of 0.0941 ° for the AP and -0.0024 "in the separation, which would give us for the present time an AP of 41.9 ° and a separation of 1,053" ... We to see that we get a result on the images obtained almost at the limit diffraction ... part of a series of 400 shots of the pair in 1x1 bin (Full Frame), with an integration of 60 milliseconds, plus 100 Darks the same time to calibrate subsequent integration. 1x1 bin mode, the Atik HR16 gives an image size of 1392x1040, taking each image about 3 seconds to load. To reduce download times and make more manageable the images, we obtained Subframes a 748x656 size ... As I always say ... Images with North up and East to the right ... This is an animation of some of the original images where you can see clearly the effects of turbulence. And here 
some of the best shots ... Where can guess the pair! 
reduction method from images "Lucky Imaging", is based on choosing the best shots of between 400 and subsequently measure them with the same tools that would measure Speckles images. This is the final shot of the pair obtained by stacking the top 14 images that represent 3.5% of the total image and its corresponding correlograms ... ... 
And finally ... measurements obtained -=-= A =-= 1352 - Date: 2011.214 Location: OANLBarcelona Conditions: Calm Full Moon-Wind Instrument: 254mm f/28.5 SC Chamber : Atik HR16 (pixels: 6.45 x 6.45) 0.060 sec. "/ Pixel Pitch: 0.1886 Delta Matrix: 41.61 -1.06 theta = rho = 1057 If we compare the measurements obtained with those calculated, we can see that accuracy is remarkable, though possibly could be refined further ... or not? What is clear is that a pair with the separation and a difference of 2 magnitudes between its components, would have been almost impossible to measure using the methods and exposure times commonly used to separate double over. My conclusions about this first experience are: - There are Further work to refine the technical and exposure times in the shooting. - In an attempt to define Speckles in the images would have to work at higher resolution as it is necessary to extend the focal length either with another Barlow taking either eyepiece projection imaging. - The results obtained with this pair encourage even try to measure pairs of magnitudes weaker than the 11.5 - should try to get a final image with better signal / noise. Since the percentage of usable images is only 3.5% in coming times will increase the number of shots. - It is clear that the methodology used "Superspeckles" will allow us splitting pairs, hitherto inaccessible, in the not too distant future. Anyway, here's chronicle my experience ... No hesitate to make any suggestion or correction to believe appropriate. A big hug 
0 comments:
Post a Comment