Over two years have past since the last image was posted. Two issues have been at play:
image quality, and
automated imaging.
Image Quality:
As other astroimagers have found, there is an inherent paradox in what we do: the better one gets, the worse the outcome. What is gained with experience is a more critical eye--since images are comprised of lengthy time exposures, and the Earth is in the habit of rotating, blurred (oblong) stars in astroimages, at first overlooked, become the nemesis. The tracking mechanism in the telescope mount that moves the scope and camera in synchrony with the earth's rotation (and thus rendering still the object being imaged) is a metal part with manufacturing tolerances (i.e. error tolerance) of approximately the width of a human hair. More error than that, and oblong stars are the result. This mount element has been replaced by a high quality custom part provided by Ovision (get the subtlety in the name?), located in France. Tracking error has been thus significantly reduced and is now within acceptable limits.
Automated Imaging:
There are literally dozens of ministeps as one travels from Point A (locating an object to image) to Point B (imaging the object). Little by little, increased intervening complexity has been layered in. Each element was fine tuned. Here's a verbal schematic of the steps involved in taking astrophotographs:
- Select an object to image
- move scope to the object
- take a test shot
- compare the test shot to a skychart to make sure the object is centered
- correct the position of the mount
- take another test shot>iterate this sequence until the object is in fact centered
- focus the telescope using a focus motor
- activate the filterwheel to choose a special filter which selectively passes light of specific wavelength of interest
- have a second scope track a nearby star and send tracking corrections to the mount which itself is already tracking at the rotation speed of the earth
- stop after 30 minutes and download the image
- repeat the cycle for a specified number of images
- move to the next object
- repeat all of the above
- stop at dawn
Needless to say, the above can keep one up all night. The latest step, however, has been to add executive software that controls and implements ALL of the above. At this point, except for opening an closing the dome, The Robservatory is now completely Robotic.
Below is is the first, successful, completely automated image.
For this automated image, I opened the dome and pushed one computer button. Fifty five minutes after sunset, the equipment came alive, located and tracked the object, did all the filtering/focusing/tracking/imaging, and downloaded the images to the computer while I slept through the night. All of the fine tuning has paid off as well: my image keep rate started at about 20%; tracking or focusing errors used to cause the remainder to be rejected. For M97, the keep rate was virtually 100%. I'm pleased that with extensive tinkering and some relatively minor upgrades, my mid priced mount is delivering a level of precision comparable to premium mounts that typically exceed 3x the cost of mine.
The image below is of M97, the "Owl Nebula". M97 is the exploded gas shell of a very old star. The green gas is ionized hydrogen; the blue gas is ionized oxygen. Total exposure time was 18.5 hours.
(Click on image to ENLARGE):