In part 3 we'll look at what can happen when the conditions aren't right - something I have experience of, unfortunately - and use this to work out what kind of equipment gives you the best chances of getting some good photos.
Iceland, December 2012
On my short trip to Iceland in 2012, we only had one night with clear skies (and no moon), but the aurora didn't show. At least, not to the naked eye. If you looked out of the corner of your eye towards the northern horizon there was a sort of glittery cloudy feel to it, but if you looked straight at it there was nothing. However, the camera proved that there was something there. This was taken using a Canon 5D (Mk I), f=18mm, ISO 3200, 30 seconds at f/2.8.
At ISO 1600 and f/4 - the limits we're assuming for consumer-grade equipment, remember - the exposure time would have had to be 2 minutes!! But that really is a sort of worst possible case. In those circumstances I'd have been quite happy to leave the camera to take a 2-minute exposure if that was the only chance I had to get something. I'd probably have taken a load of shots and overlaid them to get some star trailing, or something like that. In these dire sorts of conditions, I think your results are limited more by your imagination and your determination than by your equipment.
Norway, March 2014
In early 2014 I went to Tromsø in Norway and again the conditions weren't good. We had one night when things looked promising, but then it clouded over just as the aurora was starting to show; and we had one night when the skies eventually cleared, and we saw an aurora, but it wasn't great.
This is from the first night: Canon 6D, f=15mm, ISO 3200, 2 seconds at f/2.8. At ISO 1600 and f/4 it would have been 8 seconds, which I think would have been workable. But you'll note that the overall illumination of the scene benefits from those clouds, which are reflecting light from some villages a couple of miles away. Without the clouds, I'd have had to use a longer exposure time or tolerate the landscape being dark and underexposed.
Observations regarding focal length
You'll have noted that all my photos were taken with ultra wide lenses with focal lengths of 15mm to 18mm on full-frame (FX) DSLRs. That's the equivalent of about 10mm on crop-sensor (DX) cameras. My personal favourite is the Canon 15mm f/2.8 fisheye, which is sadly no longer manufactured but is a super lens for the Northern Lights. (The Nikon equivalent is the Nikon 16mm f/2.8 fisheye. There are also 10mm fisheyes available for crop-sensor DSLRs.)
Could you get away with a less wide lens? Yes, you could. For example I've illustrated below how these four photos could be cropped to stimulate what you'd get with a 24mm lens (15-16mm on DX). I think they're still quite acceptable from a compositional point of view. Acceptable, but I think definitely not ideal. Until I saw a half-decent aurora in Tromsø I really had no appreciation of how much of the sky they could occupy.
Conclusions and recommendations
So, what have we learned?
- You want a really wide angle lens. Fast is good but wide is essential. My four pictures here were taken with focal lengths of 15-18mm. (On a crop-sensor DSLR that's 10-12mm.)
- If you have moonlight and a bright aurora, you'll be fine with a consumer-grade DSLR and a wide-angle lens. But if you don't ... you might not be.
But it's clear, I think, that there are definitely some conditions in which consumer-grade kit would struggle, and something like a Canon 6D / Nikon D600 (or better!) and an f/2.8 lens would cope better. At the end of the day it's for you to decide how likely you think that is, how important it is to you, and whether it's worth the money.