Digital Photography 205: Astrophotography

Brooke Clarke 2013 - 2015

Background
Mounts
Polar Alignment
Exposure Time
Cameras
Camera Settings
Software
Related
Links

Background

I've been interested in astronomy for decades (See related links below) and have a good friend who has built an observatory from the ground up.  He used to move a telescope from his garage out to the street where the view of the sky was much better.  The problem was that getting the scope polar aligned took an hour or more and that was lost time every time he wanted to image.  By (moving to a hill top and) using a permanent pier and an observatory to protect the equipment from the elements and to keep him warm while observing that lost time can be eliminated.

For a few years prior to the beginning of observatory construction I told him that the three most important factors for astrophotography were: 1) The mount, 2) The mount and 3) The mount.  This is something I learned after spend a lot of time researching how to build a C-Band satellite television receiving system when there were no systems then made for home use.  Your first thought is focused on the parabolic dish and the feed system for it, but then the larger problem is how to hold and point the dish.  In a similar way many people get all focused on the telescope, but it needs to be on a mount that's solid and accurate.

I've recently seen astrophotographs (Wiki) taken with a DSLR (Wiki) camera and a wide angle lens that are stunning and am motivated to see what I can do in my backyard.

There are many things in the night sky that can be photographed.  Individual stars in visible light to me seem boring, they are just dots of light.  But to an astronomer interested in their chemical makeup or looking at their light intensity curve vs. time to see if they have planets they are very interesting, but this is a real astronomy pursuit rather than a DSLR camera application.

The things that make great photographs are groups of stars, like the Messier Objects (Wiki) or as simple as the Milky Way (Wiki).

I'm interested in satellites that are orbiting the Earth.  One area related to that is discovering satellites rather than looking at satellites whose orbital parameters (Wiki) are already known.  One way of doing this involves photographing the night sky.

Mounts

Fixed Camera

One of the simplest ways to take astrophotographs is to just mount your camera on a tripod and make a time exposure. 

Many of the delightful things in the night sky cover a wide angle and so can not be seen in a telephoto lens.  Other things are best photographed with a telephoto lens, but these benefit from large diameter optics, i.e. from using a camera with an astronomical telescope rather than using a camera type telephoto lens. 

The problem is that, depending on the lens focal length and exposure time, the star image will "trail" instead of being a point.  See Rule of 600 below.

Az-El Mount

The Azimuth - Elevation mount is simple to make and is what's used for big astronomical telescopes as well as most visual observation computer controlled telescopes.  It's easy to set up requiring only leveling the tripod head and roughly pointing to the North.  The alignment then proceeds from there.

The problem for use with a camera is that the image filed rotates so again you get star trails.

A fix for the rotating image is to put the Az-El mount on a wedge.  The wedge is tilted so that it's exactly at right angles to a line parallel to the Earth's spin axis.  This converts the Az-El mount into a Polar mount.

Equatorial (Polar Mount) (Wiki)

These mounts have an axis of rotation that's parallel to the Earth spin axis and so allow de-spinning the Earth so that the stars appear to be fixed.  With a perfect mount you could take a single exposure lasting all night, but there's not such a thing as a perfect mount.  With any real mount there will be errors that will cause the star image to smear and so there will be exposure time limits even with these mounts.  Typically higher cost mounts can make longer exposures than cheap mounts.  See the three most important things in the Background paragraph above.

Note that images of deep space objects are only possible using long exposure times with large telescopes, you can never see these using an eyepiece.

Also, you can see things in an eyepiece using a large telescope that can not be captured in an exposure (either film or digital) because "seeing" (Wiki) limits what can be seen in a single exposure, but the eye can integrate the image.  For objective diameters less than about 12 inches seeing is not a problem.

Barn Door Mount (Wiki)

Most telescope mounts are designed so that you can track a star anywhere above the horizon and keep tracking it till it gets to the horizon.  The "Barn Door" type tracking mounts trade lower cost for limited tracking range.  But this may be a good thing if there's a maximum limit on the length of a single exposure.

Meridian Telescope (Wiki Transit Instrument)

The idea is to have a telescope with only a single axis of movement aligned North South, i.e. the local meridian.  A star will cross the local meridian when it's Right Ascension equals the local mean sidereal time.

Zenith Telescope

The Ukiah Latitude Observatory has a zenith telescope where the angle from the zenith to a star can be measured with high precision.

Photographic Zenith Telescope (Wiki: PZT)

Guiding

The idea is to use a separate camera aimed at a bright star to guide the mount so that it tracks the star field.  With film photography this was done using a guide telescope and a manual tweak on top of the mounts normal rate.  Modern systems use CCD cameras and digital feedback loops to keep the scope pointed at the same star.

If there's a guide camera as part of an astronomical CCD camera it may be necessary to rotate the camera so as to be able to point the guide camera at a bright star.  My friend would like to add this capability, but for his system that's going to cost about $10,000.

Guiding is an option for any of the moving mounts above.  But at the expense of added cost and complexity.

There's also an issue about where in the optical chain the guide camera is located.  For example if it's behind the filter wheel that's a limitation.

Commercial

Losmandy StarLapse Camera Motion System

Polar Alignment (Wiki)

There's a trade off in how much time and effort you put into getting the polar axis aligned to be parallel to the Earth's spin axis and how accurately the mount is going to work.  That's to say how long can you make a single exposure without getting star trails. 
The quickest way is to point the polar axis to the North star (assuming you're in the Northern hemisphere).
A good way is to use the drift method where each iteration takes 10 to 20 minutes, i.e. you can see it may take hours depending on how good you are at it.
The best way is to use TPoint software (Wiki) that not only corrects for mechanical problems but also uses the stars as a precision clock.  But this requires that you already have done what amounts to a drift alignment, i.e. it's only practical for a fixed pier polar mount.

Exposure Time

With film cameras all the images were a single exposure so a lot of effort was expended on capturing the weakest stars given the limits on exposure time.  This included Hypering the film stock (Wiki) to raise the effective ASA number of the film.

With digital imaging it's possible to stack images (see: Digital Photography 201 Stacking Images).  The idea is toke a number of images and combine them in a stack.  Photoshop, starting with CS4, has the ability of aligning each image in the stack so it can be combined with the other images.  Each image has it's exposure limited so as to avoid star trails or problems caused by imperfect tracking but the total exposure time is much higher than the single exposure limit.

Fixed Camera

This topic has strong implications for the rest of the topics.  The problem is that the Earth is spinning and the stars are fixed in space so depending on the focal length of the lens and the exposure time at some point the stars will "trail" instead of being just points of light.  The maximum exposure time for a fixed camera is given by the Rule of 600 formula below.

Rule of 600 for Fixed Camera

Maximum exposure time of stars in seconds to avoid trails is 600 / (Focal Length in mm).

So in order to maximize the exposure time you want as short a focal length as possible.  For the Nikon I got a Tokina 11 - 16 mm lens.  This allows an exposure lasting:
600 / 11 = 54 seconds.

Cameras

When talking about digital cameras for astrophotography most people think Canon or Nikon because they are the main DSLR cameras.
I like Nikon cameras because I started photography with an all mechanical Nikon F and have some lenses and accessories that still work on my current Nikon cameras.
But Canon cameras have an advantage for astrophotography because of how the process RAW images.  They also have an advantage in that they make a great close-up macro lens and Nikon has no parallel lens.

But, there are other cameras that may be much better.

Leica makes digital cameras that produce much higher quality images (Ken Rockwell test) than either Nikon or Canon, but they are range finder cameras which are not as easy to use for normal photography as the through the lens system of the Single Lens Reflex cameras.  But for astrophotography that may not matter?

The Sony NEX-7 camera is used in a system that stitches images to achieve an overall image that competes with Hasselbald images.
PS I've heard that there may be a problem with the NEX-7 in that it does some compression that can not be avoided that shows up in star trails.

Camera Settings

ISO number (formerly called ASA number with film) has to do with trading light sensitivity and image grain.
Mirror movement can blur the image, even when the camera is mounted on a tripod, so lock the mirror up.
Output file format (Shooting Menu).  It's best to use RAW (NEF in the case of Nikon) files rather than .jpg or other processed types.
Bits per pixel (Shooting Menu), when using RAW files you can increase the dynamic range up to 14 bits per pixel instead of being limited to 8 bits per pixel with .jpg.
The default color space (Shooting Menu) is sRGB, but by switching to AdobeRGB you can pickup more in the reds.
White Balance is another thing that needs investigation, but is not an issue in RAW files since there's no color conversion.
Long exposure noise reduction (Shooting menu) is a debatable thing on the Nikon so some experiments will need to be done.  Update:  I've read that using Nikon software this can be regulated all the way to being turned off.

Software

Photoshop

Starting with CS4 Photoshop has native image stacking capability, but it's a good idea to have a dedicated hard drive for the Photoshop scratch files.

Deep Sky Stacker

Freeware to register and stack star (not planet) images including handling of dark frames.
This is the first thing to do that combines the separate images into one .tif file.

IRIS

Freeware to process the .tif image from DSS.  Mainly concerned with the dynamic range of the image.

GIMP - Tutorial @ UC Berkeley)

Freeware to adjust the color curves.

Related

Astronomy
Astronomical Binoculars - Table of Binoculars (including "Holding Power")
CCD & Image Intensifiers
Digital Photography topics (See letter D on index web page)
Optics
Navigation - relates position and time on the Earth
Photography - Nikon
Ukiah Latitude Observatory

Links

http://bf-astro.com/wideDSLR.pdf
Catching the Light 
by Jerry Lodriguss - a good overview of DSLR astrophotography

Back to Brooke's: Alphanumeric index of web pages, Products for Sale, Astronomy (& satellites), Astronomical BinocularsCCD & Image Intensifiers, Military Information, Navigation, Home page

page created  2013.