Techniques

Polar Alignment

This is the process of aligning the polar axis of an equatorial mount so that it is parallel to the earth's axis.  This enables the mount to track, in one axis, a particular location in the sky as the earth rotates.  Without polar alignment, a star on the CCD would quickly drift off the chip.

Because I use short individual exposures, exact polar alignment is not actually required.  The drift in RA caused by Periodic Error is much worse than the drift in Dec caused by poor polar alignment.  However, since I like to leave the telescope unattended while I image over several hours, I try to polar align as well as possible before starting.

I mark the floor of my balcony with the locations of the tripod feet.  When setting up, the tripod is always placed at these marks.  I do not change the elevation angle of the mount axis between imaging sessions.  In this way, even my initial setup is always approximately polar aligned, ie. the polar axis is pointing North.

In more northern latitudes, Polaris is visible and polar alignment can be done in relation to this star.  However where I am, Polaris is just a few degrees above the horizon, at best, and is always not visible.  Therefore I use Drift Polar Alignment.  My method is based on the description of the process found at the Minor Planet Observer website.

I start by pointing the telescope straight up, at a bright star.  I centre this star on the reticle in K3CCDTools, then observe for about 5 minutes to see if the star drifts in Declination (RA drift back and forth is expected because of the mount's periodic error).  Once the star has drifted, I note the direction on the screen it has drifted.  I set K3CCDTools to take 15s long exposures, then go out to the telescope and slightly nudge the aperture of the telescope south.  This is so that I can tell whether the star has drifted north, or south.  The next long exposure that appears on the screen will show a trail that will indicate whether that nudge to the south brought the star towards or away from the centre of the reticle.  If the star moved back towards the reticle during the nudge, that indicates that the star has drifted South as a result of the polar alignment error.  This tells me that the polar axis of the mount is pointing too far East of the north pole.  I make a slight adjustment to the alignment of the mount and repeat this process untill, there is very little drift in about 15 mins.

I then point the telescope at a star near the eastern horizon and repeat the steps above, except that the adjustments are to the elevation of the polar axis.  The rules for adjustment of the polar axis are summarised in the tables below, for both Northern and Southern Hemisphere observers.

Drift polar alignment takes time - I usually spend an hour aligning carefully, but pays dividends later in the imaging session.

When I first started imaging, one of the toughest challenges was actually locating the object onto the CCD.  My CCD has 640 x 280 pixels at 5.6 microns each.  With my telescope, the resulting field of view is only 12 x 9 arcmins.  In comparison, the moon subtends an angle of 30 arcmins.

The way I locate objects is to first print out finder charts for the object using Cartes du Ciel.  I print out a large scale chart for the approximate location and to locate the nearest bright star, and at least 1 detailed chart for navigation from the computer screen (see below).

I make sure that my finder scope is aligned such that if a star is at the crosshairs, it will also certainly be on the CCD.  I then look for a naked eye star in a recognisable constellation and aim the telescope at it.  Then working at the finderscope and charts, I star-hop, moving in either RA or Dec using the slow motion knobs.  I look for a mag 7 or 8 star in the finderscope and locate it in the crosshairs.

My computer, running K3CCDTools, will be showing this star on the screen.  It will be bright and be visible in video/preview mode (ie not long exposure).  At this point I usually run a long exposure, maybe 10 - 15 secs, to confirm the star against the background of fainter stars.

The hand controller will be next to the computer (see the Equipment page).  I use that to move the telescope in either RA or Dec towards the object, taking a long exposure frequently to confirm the field of view against the detailed finder chart.  I do this very carefully so as not to lose track of where I am.  If I do lose it, I backtrack to the nearest known star.

Good star catalogs are essential in order to produce charts that have sufficient star density to enable this.  I use the Guide Star Catalog (GSC-ACT) obtained on 2 CDs from Bill Gray of Project Pluto.  It has 19 million stars down to mag 15.  I recently obtained the USNO A2.0 catalog on 11 CDs, also from Bill, which has much higher star density, 526 million stars down to mag 20.  In most cases the GSC is sufficient, except in very sparse star fields.

This is extremely critical in CCD imaging, unlike visual observation, where the eye compensates for any slight focusing error.  

I begin by racking the focuser tube to marks I have made (see photo on the Equipment page).  This gives me rough focus and ensures that I will be able to actually see the star that I am focusing on.  I choose quite a bright star - bright enough to give diffraction spikes.  Here's an example of a raw image of a bright star in focus.

Note that the spikes are well defined and extend out a long way.  In a slightly out-of-focus image the spikes will be either less defined, or will show up as a double spike.

I focus using a very light touch on the focuser knob, first racking it in one direction until I pass the point of exact focus, and then very slowly back again until I get back to the focus point.  I take time to get this right, usually up to 20 mins as I alternate between the telescope to adjust focus and the computer to check the image on the screen.

Cartes is my main star-chart software.  It is freeware by Patrick Chevalley, is available here, and has such great features and looks that I do not need another planetarium program. 

I use Cartes to print finder charts.  It shows nebula outlines as well, which is useful when looking for faint nebulae.  Here (click for larger pictures) is the large scale chart to find M83 as well as a detailed chart.

Notice the labelled mag 6.7 star left of M83 on the detailed chart.  That is the same star that is visible just left of M83 on the large scale chart.  It would be visible in my finderscope and is the star that I would look for before navigating to the galaxy using the hand controller and computer screen.

The following screen capture shows my camera settings when acquiring Deep Sky Objects.  The main thing to note is that the gain is set very high.  This is to obtain maximum sensitivity.  It does result in more noise on the individual frames, but this will processed out later.  I have not really experimented with different Brightness, Gamma, or Saturation, so perhaps that's next.