Basic astrophotography with D70

I just wanted to share with the community an astrophotographical image, I made

with my D70, a tripod and a 50 mm 1.8 lens. The main object depicted is the

Large Magellanic cloud. The total exposure time is 300 secs, split up into 20

single exposures of 15 secs each (because of earth rotation and dark current -

the latter would have allowed for exposures of 1 minute). The adding was done

using the astrometrical tasks geomap and geotran and the combination task

imcombine within the astronomial data reduction package IRAF (Image Reduction

and Analysis Facility, a complicated but powerfull behemoth of originating in

the 1970s). The necessary FITS files were created in GIMP (Decompose (not as

layer) and save as FITS). Final assembly of the colour chanels was also done

with GIMP. Hope you like it. Location of picute was btw. "Cerro Tololo

Interamerican Observatory" near La Serena (Chile), and the pictures were taken

on October 14th.<div></div>

It's nice! By the way where are you? I can not see such a clear sky in UK...

 

Again, beautiful!

Oops it's Chile, I missed it just now because of the excitement. :P Do you think a single exposure can create such a picture?

Do you think a single exposure can create such a picture?

 

Dear Leon, probably not. There are two issues, one would be the earth's rotation, which would make the stars appear as stripes in short time. Depending on the lens used (i.e. focal length) this becomes obvious after a few seconds of integration. For the lens I used, i.e. a 50 mm (times 1.5), the maximum would be about 15 secs near the poles (the LMC is at a celestial latitute (aka declination) of -70 degrees, i.e. about 20 degr north of the souther celestial pole) to about 8 secs at the cel. equator, e.g. in Orion. This can of course be compensated by using a tracking device which accounts for the earth rotation. But for non-cooled digital cameras, such as the D70 there is another limit, namely the dark current. Whenever a CCD integrates, it accumulates charge, i.e. signal. While current chips have become less prone to this, the generation of 10 years ago could be filled up completely within one or a few seconds. With a modern DSLR, you might be able to do exposures of up to about 1 minute, depending on model and ambient temperature. For exposures longer than 1 sec, one should also take a dark frame, which is a picture taken with the same exposure time but without exposure to light, this is then subtracted from the image (my D70 did this for every frame). Another limit, affecting both film and digital is the ambient skuy brightness, which depends on lunar phase and light pollution.

 

Hope that answered some of your questions.

Thanks a lot Martin.

 

Interestingly, I was always want to catch the star trail, but realize that it is quite hard of a DSLR. I DID expose as long as 300 seconds with my D50, not so bad, but the main limitation was the bad ambient lights in big city (London, sigh..).

 

I am quite impressed your idea to combine several pictures. Because image registration is a very hard work - I majoyed in the 2D/3D automatic image processing. Some maunal work is essential, but it can not be perfectly done.... You result has been quite good, although there are minor double-shadow. (No offence, just tiny amount) I like it anyway!

 

I love astrophotography, here is my shot of a harvest moon last week. hope you like it.<div></div>

This is wonderful, Martin. <br><br>

If I understood correctly, IRAF has the capability of (easily) correcting for field rotation? And if I may ask, what ISO setting did you choose?

I have a photo in my gallery of a tree in the full moon exposure time was 30s, if you view larger you can see small star trails. This type of photography is really fun!

 

Good job Martin!

Dear Michael,

 

IRAF is a full blown professional astronomical software package, which can perform all kinds of things, virtually everything needed to get results from raw data as taken from the telescope - in fact it is probably the most used software by astronomers. The good news is, it can be obtained for free (www.iraf.noao.edu), the bad news is:

-it is a program from the 1970s and it behaves like one

-installation can be a pain

-only runs on UNIX based OSs, like linux (a debian package exists, which should make installation much easier), the various generic UNIXes, and MacOS. No Windows version available.

 

But once one has overcome these obstacles, IRAF can be quite useful, since it has features photo manipulation programs (such as GIMP,PS,...) usually lack, such as the centroiding of points, such as stars (one of the most (important tasks in astronomy) and the registration of images (i.e. making them fit). Since I've been doing that for years on astronomical data, I decided to try it on my digital images from the D70, and it worked - was a lot of work though. The aformentioned task "geotran" basically is the heart of the whole process. It calculates a transform of an image to a reference image (the trick is to choose one of the images, usually from the middle of the sequence as reference and to register the others to this image), which is accurate to (in my case here) less than 0.1 pixel (it accomodates for nearly everything, i.e. shifts, rotation, distortion, etc.). As an input geotran needs a list of (at least 5) coordinates of the reference image as well as a list of coordinates of the same objects for the image to be transformed. The task geotran then actually performs the transformation of the image. The coordinates of objects like stars can be derived by using a simple IRAF task like imexamine.

 

Short answer: yes Iraf is capable of doing that.