Adventure in Astrophotography

Berry Summit SM16-40-3

Astrophotography, or the photography of nighttime skies, has become fairly popular in recent years, thanks largely to improving sensor technology. It’s still a fairly challenging type of image making, because even in areas destitute of the horrors of light pollution, stars aren’t very bright, and require long exposures and high ISOs and fast apertures. To get an idea how feint the stars can be, consider the following image, shot with a 10mm lens at f2.8, ISO 800, 30 second exposure:

Berry Summit SM16-43

The foreground is brighter because I shined a headlamp on it. The stars are quite faint. Even worse, because of the 30 second exposure, some slight motion is detected in the stars (yes, the stars “move” — although, of course, it’s the motion of the earth that’s the prime culprit!). Here’s a close-up of the stars which shows just a bit of movement:

Berry Summit SM16-43-2

The width of the focal length affects how much motion will show up in the image. It shows up only slightly in this image because of the use of an ultra-wide angle lens. Here’s the amount of motion captured by a 35mm lens (on an APS-C camera) with an 86 second exposure:

Berry Summit SM16-38

To get around the problems created by motion of the stars, I have attempted to make use of a technology introduced by Pentax — something called the “astrotracer” function. As Pentax explains:

When mounted on [the appropriate Pentax] body, the O-GPS1 provides the advanced ASTROTRACER function, which couples the unit with the camera's SR (Shake Reduction) system for the effortless tracing and photographing of celestial bodies. The unit calculates the movement of stars and other objects using the latitude obtained from GPS data and the camera's alignment data (horizontal and vertical inclinations and direction) obtained from the unit's magnetic and acceleration sensors. It then shifts the camera's image sensor in synchronization with the movement of these objects. This means that all celestial bodies are captured as solid points rather than blurry streaks, even during extended exposures. It also simplifies the astronomical photography setup, since it requires only a tripod and eliminates the need for an additional accessory such as an equatorial telescope.

The requisite O-GPS1 accessory basically provides GPS capability to Pentax DSLRs. Typical of such units, it fits in to the camera’s flash shoe. Here’s what it looks like:

Screen Shot 2016-09-05 at 10.22.15 AM

While this technology is currently confined to Pentax DSLRs (essentially K-5 and later), there is hope for other brands. All you need is to combine a camera body with IBIS (in-body-image-stabilization) and GPS capability. Any camera systems that makes use of IBIS (Sony, Olympus and Panasonic) could, if they so pleased, add astrotracer functionality to their cameras, either through an accessory, like the O-GPS1, or even through GPS capability built into the camera (like with the Pentax K-3ii or Pentax K-1). Camera systems without IBIS (mainly Nikon and Canon), will not be able to develop this technology — but they, no doubt, have other advantages to make up for this lack.

How well does the astrotracer technology work? That’s what I wanted to find out when I took an evening trip to Berry Summit, about 35 miles outside of Eureka on Highway 299.

In order to use the O-GPS1 astrotracer functionality, you need to “calibrate” the unit. This involves turning the camera in a circular motion in three separate directions, as illustrated in this diagram:


There are two levels of calibration: normal and “precise.” Normal calibration is fairly easy to attain. Indeed, the camera will often signal it has been successfully calibrated after only moving it in two directions. Precise calibration is more difficult, but preferable.

The O-GPS1 allows one to take images as long as five minutes. So I started with a five minute exposure shot with a 35mm lens:

Berry Summit SM16-40-2

As can be seen, at five minutes, there’s slight motion in the stars. So going forward, I used exposures between 60 seconds and 140 seconds. Here’s an image shot with a 15mm lens at 139 seconds:

Berry Summit SM16-44

At over two minutes, we can see slight motion in stars. What about at 65 seconds?

Berry Summit SM16-45

The stars here show virtually no motion (and what motion that does seem to exist may be caused by lens coma).

Here’s one more close-up example, this time shot at 96 seconds with a 21mm lens:

Berry Summit SM16-42

From these examples, I would suggest that keeping exposures under 100 seconds, at least with wide angle lenses, seems to be the best policy.

One of the drawbacks of moving the sensor to capture images of the stars is that it will cause blurring in any terrestrial objects that one might wish to include in the shot. Here’s an example of what happens to the horizon during a five minute shot with the astrotracer functionality enabled:

Berry Summit SM16-40

The horizon is blurry. There are a number of ways to get around this. One might be to use a brief light source, such as a flash. This fence in this image was illuminated by the headlights of a passing car:

Berry Summit SM16-46

This image is from a 139 minute exposure. Look at the fence up-close:

Berry Summit SM16-46

Keep in mind that this is an image shot at ISO 800, so there’s some noise reduction being used which is making the image look a little softer than it really is. But this is not bad at all.

The cleaner and more safer method for including terrestrial foregrounds in astrophotography images is to combine shots, one of the night sky using the astrotracer functionality, the other of the foreground with astrotracer disabled. Here’s a combined shot of two images shot with a 10mm lens:

Berry Summit SM16-43-Edit

And, to include, here’s a combined shot, with the foreground shot with a 15mm lens, and the sky shot with a 21mm lens:

Berry Summit SM16-42-Edit