Star Trails and DSLRs

I am going to New Mexico next week for Spring Break and while I am there I want to try to go somewhere really dark one night and photograph star trails. However, I am running into a technical problem.

My equipment is a Nikon D80. I took this test photograph with it the other night in my driveway. Notice the purple glow in the upper corners. After some research, it appears that this is heat.

Apparently, most digital cameras are ill-suited for astrophotography because at very long exposures the heat that is generated by the circuitry manifests itself on the CCD as a purple glow. Supposedly, the are DSLRs that are specially made for astrophotography that are cryogenically cooled, but that is way too much for an amateur like me.

My idea for a poor man’s fix, is to get one of those first aid chemical cold packs, wrap it in a towel (to prevent condensation), and strap it to the back of the camera. Thoughts? Has anyone tried this? Are there other methods that I can employ to work around this issue? Any and all suggestions, comments, and tips are appreciated.

Easy thing to do is probably to just center the image better and crop out the purple corners.

Well the problem, as I understand it, is this: That was approximately a one hour exposure. I want to do a 6-7 hour exposure, which at that length of time will saturate the CCD with the purple glow.

Why not take 6 one hour exposures, and overlay them with your photo software?
That should get you the full trails, without the purple.

Hmmm, but how long would the CCD need to cool down between shots?

Main problem I see with ice packs is condensation on the inside surface of the camera - a towel would only help with keeping the outside dry. Aside from that, they’re not going to be cold for six hours - would you be able to change them without jiggling the camera?

Sometimes, good old film is still the best option. :smiley:

I’d be very careful with forced cooling. Condensation is a big concern, not just on the cold pack but also the cooled camera.

You might try taking a “dark frame” by putting the lens cap on and taking an image with the same exposure time, at the same temperature (or as close as you can manage). Then subtract it from the original image. This isn’t perfect because the temperature would shift between the two images, but it’s better than nothing.

Maybe you can take, say, one 10-min dark frame, then six 10-min exposure images, then another 10-min dark frame. Sum the 6 images, then subtract the sum of the dark frames multiplied by 3. That way you at least compensate for a linear change in dark current.

No, you’d end up adding the purple glow as well.

You could take a one-hour exposure, or whatever, and try Photoshopping the trails based on that.
Note to self: don’t sell that 35mm SLR body in the closet…

Noise is additive here, not signal. So at 6 hours the purple will likely overwhelm all those dim startrails. It shouldn’t be too hard to subtract an hour’s worth of purple noise from each image before adding them.

Well, I went out in the desert last week and played around with night shots. By and large, I am pleased with the results. As expected, I had the purple glow from the heat manifestation, but a quick crop took care of that.

Also, as dark as it was, I was not dark enough. Notice that the sky in the lower half of the photo is not as dark as the the upper half. I assume that is light pollution from T or C, NM. All in all though, I think it turned out well for a first serious attempt at photographing star trails.

Let me know what you think:

Original photo

Cropped photo

One thing I forgot to mention: Other than the crop, I haven’t tweaked or modifided the photo at all. What you see is what came off of the camera.

Nice color on those trails!
I’m surprised you didn’t pick up a north/south spy satellite though.
That was about a 1/2 hour exposure?

Thanks! :slight_smile:

Yep. Here is the exif data, if you’re interested:

I took some with longer exposures, but while I caputured some excellent trails, the lower portion of the sky became increasingly muddy due to light.

In the future, apart from finding an even darker spot from which to photo, I think I will step the aperture down a step or two. I’m thinking that will eliminate the very faint trails and give a “cleaner” looking photo.

Looks good, but there’s some camera shake. Next time you might want to try shortening the tripod legs all the way, and/or choosing a more sturdy tripod. It may also help to choose a location shielded from wind.

Cool. Pardon my astronomical ignorance, but why are the stars going in a circle? Shouldn’t they just go from one side to the other? :slight_smile:

The reason that they appear to be traveling in a circle is because I aimed the camera at Polaris (North Star) and as the Earth rotated (the camera and everything else on it), the stars appear to circle Polaris (which remains stationary). Bear in mind that I am not an astronomer, and I am sure that someone will be along shortly with a more eloquent answer.

So does the fraction of a circle a star trail in the photo makes equal the fraction of 24 hours the shutter was open? : )

Not exactly, but almost. It’s the fraction of a sidereal day that the shutter was open. A sidereal day is about 23 hours 56 minutes.

Again, not my forte, but I assume so. I imagine that if you were to go north of the Arctic Circle on December 21, and left the shutter open for 24 hours, you would get complete circles.

As I said, 23 hours and 56 minutes would be enough for a complete circle, in either of the polar regions in the middle of winter.

An idea you might try to get rid of (or rather mask it) the unwanted light area, would be to set up your shot in such a fashion that there’s a tree, or rock formation, or some other feature, in the foreground. Then briefly illuminate that feature with a simple flashlight during the exposure.