Getting there. Staying there. Looking there..

How I do deep sky astronomy imaging with a modified webcam.

I have been imaging deep sky objects with webcams since Jan 2003. My equipment and methods are constantly refined.

My current deep sky imaging setup is listed here

This page details how I typically use my setup to image.

1. Remove observatory roof, and any sides required to see target object.
My observatory has 6 foot walls. Each wall is divided in two, and the upper half of each wall can be removed to allow viewing in a particular direction. The more of the removable sides that are left in place, the better the shelter from light and wind. It is possible to leave the roof on when observing near the horizon, provided 3 sides are left up to support it.

2. Plug in the power. Switch on the scope.
Wait for the comforting beep from the Autostar controller in the house to confirm the mount is alive.
I used to use a rechargeable battery to run my scope. It worked well for a year, and then suddenly stopped working. Not good. So I wired my mount up to the 12 volt output of an unused computer power supply. This has proved flawless, and also povides other voltages for fans, peltiers, etc.

3. Remove dust caps and attach dew shield.
I have a long dew shield made from a camping roll mat and duct tape.
I cannot leave the dew shield attached, otherwise I would need a very tall observatory indeed!

4. Visually inspect everything. No tangled wires, dead rodents, cables unplugged etc.
Cobwebs are also sometimes a problem! Also check inside of the OTA to make sure nothing creepy or crawly has made its home there.

5. Move back to the control room.
If its cold, put the heater on and make coffee.

6. Startup the finderscope camera.
I have an unmodded Toucam inserted in the eyepeice of my finderscope. This allows me to "look" up the finderscope from the control room.
There is also another toucam, this one modified, which acts as a guidescope. Both these cameras are plugged into a USB hub in the observatory. The hub is linked to the workstation by a passive 5m usb extender.
The USB cable does not have enough bandwidth to run both cameras at the same time. It will allow both cameras to be plugged in - but I have to use Computer Management to switch off the one I am not using. This is a little irksome, but much less hassle than running outside each time I wish to swap cameras.
I startup k3ccdtools2 and switch to WDM mode. I adjust the settings to get max exposure and gain.
As my telescope is "parked" pointing at the North Celestial pole.. Polaris will be on the screen.
If its not, something is very wrong!
My mount is controlled from the Autostar Handbox. I can steer NSEW at a variety of speeds from guide rate to 3 degrees per second.
At this point I will check the controls are responding, and I can move Polaris about in the finderscope.

7. Plug the main imaging camera into the laptop
A third toucam is mounted at prime focus of the main telescope. This is a black and white SC3 toucam with an ICX424 chip.
I get k3v2 fired up. Load the factory defaults to get true raw mode. Make sure everything is working.
I start the imaging camera running at max gain and exposure, but there is nothing to see yet.
Sometimes I might take a flat field.
I do this by hanging a white pillowcase over the OTA, and directing an external floodlight at the telescope.
Quickly processing this AVI and performing a histrogram stetch will quickly show if the CCD chip needs a wash.

8. I then run my internet clock software to confirm the correct time. Autostar likes to have this information before doing anything. I type it into the hand controller along with the date.
As my scope is parked from the previous session, it remembers its alignment, and only needs the correct date/time to function. Well, thats the theory. Sometimes it forgets and I have to do a 2 star alignment. This is done from the control rom using the ideas in point 10 below.

9. Now I fire up Skymap on the workstation and plug the RS232 cable into autostar and get skymap to connect to the handset.
I do not leave the handset connected to the computer unless I need to, as I sometimes think strange things come out of the COM port and make Autostar do odd things!
Skymap will jump to "real time" mode, and the little cross that makes the scopes position should be on the NCP.

10. I now test everthing is working by slewing to a bright star. Perhaps Vega. Using Skymap to select targets is much easier than navigating Autostars's tedious menu structure.
If all is well Vega will appear somewhere on the finderscope screen. On this computer screen is a small piece of bluetak.
This is the "camera position". I steer the mount until the bright star is under the blob.
It should now appear on the main imaging laptop screen.
Sometimes its does not! The blob of bluetack needs moving, perhaps the finderscope has moved slightly since the last session, or I have moved the camera slightly.
So I tell Autostar to do a very tight spiral search until I get the star on the main imaging laptop. I fool Autostar into doing this by telling it that the telescope has a focal length of 8000mm and selecting the spiral search mode.
Then I adjust the bluetak blob to the correct position on the finderscope screen. It normally does not need moving all night.
Another approach is to defocus the star a lot. A big defocussed star is easier to find than a small focused one.

11. Now I check my collimation. I am still working on my collimation methods, so this is not part of my standard imaging process yet.
But I do defocuss the bright star using my electric focuser. I check this looks reasonably collimated.
If it is not, then I do some collimation!
I will frequently check collimation with a laser collimator before imaging.
You can find some of my own collimation suggestions here.

12. Next I focus on this bright star. I use a second or two of long exposure and keep adjusting the electric focuser controls until the diffraction spikes are needle points.

13. Updating pointing
Autostar currently thinks its pointing at my bright star (Vega). But I had to move it a little after the GOTO to get the star on the camera.
LXD55 GOTO is not a Losmandy Gemini GOTO!
Therefore I use skymap to tell autostar to sync itself to Vega. This updates its little brain with better pointing information.

14. GOTO target.
I open skymap to show me a large portion of the sky around my target.
I then use skymap to slew to the brightest star in this area.
I use the finderscope camera and bluetak blob to get this new star on the screen and centred.
I then sync to this star.

I repeat this a few times, getting closer and closer, until I finally GOTO my target itself.
I prefer this "walking in" method to the automated version in Autostar - high precision mode.
Often HP pointing will pick a sync star on one side of the meridian and the target on the other side.
Lots and Lots of wasted slewing to cross the meridian!

15. Frame target.
Now I start K3v2 doing some long exposures with the main imaging camera. I use the shortest expsoure I can use and still identify the target position on the screen.
Sometimes its not on the screen!
I have to use skymap to identify the star field and move the scope accordingly.
I then move the mount controls very slightly to frame the target how I want it.
Then I sync to the target!

16. Now I run long exposures, usually about 15 seconds. Gain at maximum. Whatever settings I feel best for a particular target.
Sometimes more sometimes less. I check and double check the record button is pressed in!

17. Staying there.
Although my mount is rather nifty for finding things, it does have periodic error.
There is electronic PEC, but I like to guide anyway.
I shut down K3v2 and switch off the the finderscope camera in computer management. I switch on the guidescope camera.
I open GuideDog.
I tell Skymap to relinqiush control of the scope.
I start Guidedog, connect to the Autostar handbox and start doing 1 second exposures.
I keep increasing until I see a star on the screen! If I don't get a guidestar with a 5 second exposure, I swear a bit.
I go outside and move the guidescope very slightly on its adjustable mount. Maybe I have to focus it a touch. I keep moving it about a bit, looking through the kitchen window at the screen until I see a star!
Sometimes it takes 5 seconds exposure to see the guide star. Other times I have one that shows in 1 or 2 seconds.
The brighter the better. It is worth spending time hunting for a bright guidestar that will give good guiding.
Good tracking gives better pictures in the end!
I then tell guidedog to lock onto this star.
Then I check the East-West alignment now. Sometimes I need to tell GuideDog to reverse E-W. I think this depends what side of the meridian I am on.

18. Image!
Now I leave everything running for perhaps an hour to gather exposures. More exposures the better!
I like to have at least 100 raw frames to select from on a deep sky object.

19. Mosaic
Many deep sky objects are large! Larger than my FOV.
THerefore I will often move the frame about a bit during the hour to end up with a final image with a wider field of view.
Not always. But sometimes.

20. RGB
If I want a colour picture I must take a series of exposures using my red, green and blue filters. I do not have a filter wheel yet. I have to go outside, remove the camera from the scope, unscrew the IR filter and screw on the colour filter.
Then I take a series of exposures in each colour. Normally 5s longer than the L layer exposures.
I always make sure I do my exposures in order: Red, green then blue. As soon as possible I rename the AVI files to a meaningful name. Later on it would be very hard to tell which AVI is which.

21. Star bloat
Some fields of view will have a number of bright stars. With very long exposures these stars get very bloated in size. So my final exposure will be without a colour filter, but of a very short duration - perhaps one third of the length. This AVI can be used later to control the star sizes when processing.

22. Stopping
Hard to know when to stop imaging a particular target.
But between the 80 frames and 150 frame mark (plus RGB if required), I will stop and choose another target.
Normally one nearby. The previous AVI will be copied to the computer in my study (warm!) and I will start processing this one whilst exposing the next one.
Between targets I might check the secondary mirror for dew/frost.
If its looking bad, I dry it with a hairdrier.
You have no idea how much noise a hairdrier makes in the garden at 2am!!!

23. Stumps
The end of the session is dictated by a mixture of
b)Need to get up in the morning
c)Too tired.
e)Something broken
I will tell Autostar to park the scope. This slews to OTA to the NCP.
I will take a sponge outside and clear the worst of the dew of the exterior of the equipment.
Remove dew shields.
Replace dust caps.
Replace removable sides.
Unplug the scope. Once I left it plugged in for 18 hours and found it crashed ito the side of the shed. Happily no damage.
Put roof on.
Then I come inside and take some dark frames on the imaging laptop.
Finally I unplug all the cameras and turn screens off.

22. Post Processing.
First pour a large glass of Malt Scotch to counteract all the coffee I have been drinking, and continue to process my images.
Some of them I will post to QCUIAG Yahoo Group. 

Page last updated 2005-09-15