Re: Obs ML (4352), 8 Sept 2005: USA 129

From: Marco Langbroek (marco.langbroek@wanadoo.nl)
Date: Fri Sep 09 2005 - 08:46:19 EDT

  • Next message: Marco Langbroek: "Re (2): Obs ML (4352), 8 Sept 2005: USA 129"

    Björn Gimle wrote:
    >>
    >> Method: Canon Digital Ixus 400 + ASTRORECORD astrometric software.
    >> Camera at manual mode, "10 second" (= 10.7 second) exposure, wide field
    >> f2.8/7.41 mm.
    >>
    > Does this software consider "wide" angle pillow distortion?
    
    ASTRORECORD uses the Turner method and considers the common distortions of a 
    wide field lens. It originally was designed for professional quality reduction 
    of "small camera" (35 mm analogue SLR) meteor multistation photographs for 
    orbital determinations. Results on meteor trajectories and orbits obtained using 
    this software have been published in several peer-reviewed papers.
    
    These meteor images are taken typically with an analogue 35 mm SLR with 35-50 mm 
    lens, hence "wide" angle. The images of my digital camera are equivalent to a 43 
    mm lens on an analogue SLR, so comparable (the field width is some 50 degrees).
    
    > How many star positions do you, Marco,  use in one reduction?
    
    Typically some 100 stars. Good results actually can be obtained with much less 
    reference stars. Important is that they spread well around the trail.
    
    > How far from the satellite position?
    
    I use stars spread over the whole image, taking care to make sure the trail is 
    totally surrounded. I als try to have the immedate vicinity of the trail well 
    covered.
    
    > Have you verified the accuracy by taking different stars?
    
    The software calculates and gives feedback on the deviations in the average 
    positions measured from star to star. With my camera setup, the internal 
    accuracy of measured star positions is typically given as 25" to 30" provided 
    enough stars are measured. The accuracy of the satellite positions is less than 
    this, as this is also influenced by the timing error, and how well the trail 
    shows up on the image (how well the trail ends can be determined).
    
    There are some tests of accuracy relative to other stations, and a test where 
    positions obtained on ISS are compared to an ISS TLE, on my website:
    
    http://home.wanadoo.nl/marco.langbroek/satcam.html
    
    This also discusses some pittfals encountered.
    
    > I have only tried software that assumes flat-field, narrow (CCD) images.
    
    AstroRecord was specifically designed for wide field images with field sizes of 
    several tens of degrees, see above. It does not assume flat-field.
    
    The whole setup here is an experiment, to see whether it works for (bright) 
    satellites. After a month of operation, is seems it does work, bar a few 
    pittfals. Feedback is welcome, as I strive to continuously improve.
    
    - Marco   :-)
    
    -----
    Dr Marco Langbroek  -  SatTrackCam Leiden, Cospar 4352
    Leiden, the Netherlands, 52.15894 N, 4.48883 E (WGS84), +5 m ASL
    
    SatTrackCam: http://home.wanadoo.nl/marco.langbroek/satcam.html
    Station (b)log: http://sattrackcam.blogspot.com
    e-mail: meteorites@dmsweb.org
    -----
    
    
    -------------------------------------------------------------------------
    Subscribe/Unsubscribe info, Frequently Asked Questions, SeeSat-L archive:  
    http://www.satobs.org/seesat/seesatindex.html
    



    This archive was generated by hypermail 2b29 : Fri Sep 09 2005 - 08:53:02 EDT