Differences

This shows you the differences between two versions of the page.

--- howto:astrometry [2009/04/30 14:34] – pixel size formula, small corrections mates
+++ howto:astrometry [2009/11/29 14:30] – wildi
@@ Line 1: / Line 1: @@
 ====== Image astrometry parameters ======
-RTS2 is able to support on-line astrometry, with corrections being fed back to the telescope as correction parametes, thus significantly improving the telescope pointing. Following is list of parameters which affect astrometry, and a guide how to set them up.
+RTS2 is able to support on-line astrometry, with corrections being fed back to the telescope as correction parameters, thus significantly improving the telescope pointing. Following is list of parameters which affect astrometry, and a guide how to set them up. These parameters are arguments to the rts2-camd-* device drivers and are best set in the file  ///etc/rts2/devices//.
-  * **XPLATE** and **YPLATE** - size of pixel in arcseconds in X and Y direction. This is specified by --plate <x>:<y> parameter, passed to camera daemon (preferably in ///etc/rts2/devices// file). If you have 0.3 arcsec in X and 0.2 arcsec inY (=non-square pixel size), you specify //--plate// 0.3:0.2 (//not supported on the astrometry side, though :-|//). **XPLATE** and **YPLATE** are normal variables, displayed in **rts2-mon**. Camera daemon change them when you change binning - their change should be immediately visible. You can also change them directly in **rts2-mon** - but those changes will not be recorded. **XPLATE** and **YPLATE** shows only when //--plate// parameter is present. Precision needed for the astrometry to run smoothly is relatively high (~5%), so it is best to get the values from dry runs of the astrometric routine. Crude estimate may be done in various ways, pixel_size[arcsec]=pixel_size[um]*206.265/focal_length[mm] being a quite useful formula. A good way is also getting some known field (with a nebula or a small galaxy) and trial-error with the astrometric routine.
-  * **FLIP** - that depends on readout direction and used optics. Can be also regarded as indication of negative XPLATE - e.g. arcseconds rose in different direction along X axis then is common. It is 1 for most cameras, the only exception so far is Andor, which have 0 flip for EM channel and 1 for non-EM channel. This is set in the driver and unless you put extra mirror in optics, should be correct. If needed, you can set flip with //--flip// parameter passed to camera daemon. Again, **FLIP** is visible in **rts2-mon**, and can be changed from **rts2-mon**.
+  * **XPLATE** and **YPLATE** - size of pixel in arcseconds in X and Y direction. This is specified by //--plate <x>:<y>// parameter, passed to camera daemon. If you have 0.3 arcsec in X and 0.2 arcsec in Y (=non-square pixel size), you specify //--plate// 0.3:0.2 (//not supported on the astrometry side, though :-|//). **XPLATE** and **YPLATE** are normal variables, displayed in **rts2-mon**. Camera daemon change them when you change binning - their change should be immediately visible. You can also change them directly in **rts2-mon** - but those changes will not be recorded (if you restart camera daemon, they will be lost). **XPLATE** and **YPLATE** are present only when //--plate// parameter is included among camera daemon parameters. Precision needed for the astrometry to run smoothly is relatively high (~5%), so it is best to get the values from dry runs of the astrometric routine. Crude estimate may be done in various ways, pixel_size[arcsec]=pixel_size[um]*206.265/focal_length[mm] being a quite useful formula. A good way is also getting some known field (with a nebula or a small galaxy) and trial-error with the astrometric routine.
-  * **ROTANG**, which is an angle of north direction from vector Y axis. It is 0 when north is up on images, 90 when north is towards left (and **FLIP** is 1). **ROTANG** changes on german equatorial mount (GEM) by 180 degrees if you cross pole and go to the opposite side of the sky - eg. from hour angle 2 to hour angle 14. Default mount **ROTANG** is specified by //-r// parameter, and shows as **MNT_ROTA** in **rts2-mon**. You can also edit it there. **MNT_ROTA** changes on GEM by 180 degrees automatically. Please be aware, that **ROTANG** written to the FITS file can be sum of different variables, all having **RTS2_DT_ROTANG** flag set when creating variable. RTS2 automatically sums all variables having **RTS2_DT_ROTANG** to form resulting **ROTANG**, which is then written to FITS headers. Check your mount source code for details - e.g. for OpenTPL mount with derotator, derotator angle is added by this code:
+  * **FLIP** - that depends on readout direction and used optics. Can be also regarded as indication of negative **XPLATE** - e.g. arcseconds rose in different direction along X axis then is common. It is 1 for most cameras, the only exception so far is Andor, which have 0 flip for EM channel and 1 for non-EM channel. This is set in the driver and unless you put extra mirror in optics, should be correct. If needed, you can set flip with //--flip// parameter passed to camera daemon. Again, **FLIP** is visible in **rts2-mon**, and can be changed from **rts2-mon**.
+  * **ROTANG**, which is an angle of north direction from vector Y axis. It is 0 when north is up on images, 90 when north is towards left (and **FLIP** is 1). **ROTANG** changes on german equatorial mount (GEM) by 180 degrees if you cross pole and go to the opposite side of the sky - eg. from hour angle 2 to hour angle 14. Default mount **ROTANG** is specified by //-r// parameter, and shows as **MNT_ROTA** in **rts2-mon**. You can also edit it there. **MNT_ROTA** changes on GEM by 180 degrees automatically. Please be aware, that **ROTANG** written to the FITS file can be sum of different variables, all having **RTS2_DT_ROTANG** flag set when the variable is created. RTS2 sums all variables with **RTS2_DT_ROTANG** flag set to form resulting **ROTANG**, which is then written to FITS headers. Check your mount source code for details - e.g. for OpenTPL mount with derotator, derotator angle is added by this code:
 <code c++>
         // nasmith derotator