David's Astronomy Pages Notes - Session 306 (2008-09-27) Notes (S305)   Notes (Main)   Home Page   Notes (S310)

Goto Images from 2008-09-27

Automated Slew, Locate and Guide

Newly developed procedures for automatically initiating auto-guiding (autoguide between frames) after a slew/target locate was successfully tested during session S306. Previous autoguiding was initiated manually and switched off manually.

As before user-written routines for identifying and following the brightest star in each image are used with guide (nudge) between frames to keep the frame positioned on the same star-field within 1-2 pixels  (2-5 arc secs at my image scale). The routine is robust to passing cloud etc, whereas the standard autoguiding through CCDSoft is not and can lead to the scope trying to guide on noise in 'real world' Scottish conditions. 

The results of the tests appear very good, and will allow more automated acquisition of quality time series, for a number of applications requiring long sequences of images : variable star time series (flare stars, eclipsing binaries), exoplanet transits or asteroid/comet tracking. 

The purposes of auto-guiding between frames are 4- fold

• The target object remains appropriately centered and doesn't wander to one side

• Stacked images / animations will be able to use the full AOI of the image 
      (if there is uncorrected star drift then the area of full frame coverage is successively 
      reduced with time, black panels appear on the sides of stacked images reducing 
      the 'live' area and necessitating cropping for presentation)
  

• Comparison or check stars located near the edge of the image area, will remain in view for the whole session
      (if comparison or check stars drift near the edge or out of the frame then photometric analysis will not be possible
      manual intervention/re-centering is not always convenient and can produce a time gap in the data)

• Variable Star and Comparison & check stars remain positioned on the same part of the image through the session leading to more accurate photometric data
      (if variable and comparison stars drift across the frame through the session, the results of photometric 
      analysis are much more affected by the quality of the flat frame and any uncorrected vignetting across the star field. 
      Previous analysis has shown that it is difficult to produce perfect flat frames and even then they may not correct
      light variation across the frame if there is sag in the imaging system (due to a heavy camera !)
      Keeping stars located on or near identical pixels through the run should reduce the impact of uneven lighting.

An analysis of the automated guiding urposes of auto-guid

Back to Top

Analysis of automated guiding performance - SS Cyg

Automated guiding between frames was active during a time series run from SS Cyg, involving run of 120 frames (30s exposures) over a 82 minute period. 

Routines automatically selected the brightest star in the image as the guide star (GSC 3196:701, Mag +8.5) 
An position error threshold of 0.5 pixels relative to the first image was used to make trigger guide corrections.

Analysis of 120 frames from SS Cyg over a 82 minute run showed that the average pointing error from the mean star position was 0.74 pixels   This is equivalent to 1.8 arc secs. This contrasts with an average FWHM of 3.5 arc secs over the run. 

Based on overall drift rate between frames when guide corrections were not triggered,  then the amount of star drift across the 82 minute run would have been 58 pixels, equivalent to 2.4 arc minutes.  This would have significantly compromised the ability to carry out an automated long time series if no guide correction was employed.

Qualitative Analysis of autoguiding performance

Image below shows simple stacking of 120 frames from a 1h 22m duration run of SS Cyg No alignment of images has been performed.  Apart from slight bloating of stars there is clearly no star drift (Unguided there would be noticable star drift over an interval of 82 minutes)

Stacked CCD Image (Non-Aligned) 120 x 30s exposure (average combine), 2x2 binning, V filter,   2008-09-27 23:53h UT (#306232-351)

Image below shows aligned image stacking of the 120 frames  Alignment of images has been performed.  The stars are sharper in this image compared to the non-aligned image above. but the difference is not that great and shows the obvious success  of autoguiding between frames

Stacked CCD Image (Aligned before stacking) Image details as above

Quantitative Analysis of automated guiding performance

Back to Top

Automated Slew & Jog Positioning Performance (S306) 

Analysis of data from session S306 shows that Pointing Error after slews was consistently less than 4 arc minutes (P50 2.1 arc mins, mean 2.2 arc mins), with final pointing error after jog positioning consistently less than 0.6 arc minutes (P50 0.4 arc mins).

See earlier analysis of Slew and Jog Positioning performance from session S297-S305

Graph showing analysis of Automated Slew/Locate Positioning performance (2008-09-27, S306)

Back to Top