I gave a talk at the AAIC last month about Nightscapes.
A common question that comes up is what exposure length F ratio is best?
My suggestion is to use the dxomark.com sensor ISO rating for your camera as they test what is the highest ISO that camera does before image degradation sets in.
For example with a Nikon D800 the Nikon engineers have done various optimisations so that there is no gain above ISO1600 and probably very little below.
There is the practical aspect of being able to see the image once its downloaded and if you can't it could lead to errors when doing a mosaic (not overlapping, not noticing a cloud affected image etc).
But bringing over the usual dedicated astrophotography camera exposure approach to DSLR makes a lot of sense. In dedicated astro cameras there is very little gain applied (unlike DSLRs where ISO's above a certain point are amplyifying the signal perhaps many times, this would be similar to stretching a CCD image in preprocessing).
You have to build the signal up to get above the noise floor of the camera. Longer exposures give better images, that take more processing and are more robust and they show deeper colours and more detail.
But not everyone wants to lug around a good quality GEM and autoguider to do that and it would prevent getting a landscape in there as well.
So limiting exposures to what can be achieved with something lightweight and very portable like a Vixen Polarie or similar is the go.
So I did a test the last 2 nights. The first was with the very lightweight Fuji XE1 which as far as I know has the lowest noise at high ISO of any APSc camera mirrorless or DSLR. It is also very light, has a superb white balance engine, terrific colour and has no anti aliasing filter so it tends to be a bit sharper than many competing cameras. Fuji lenses are also very high quality for much less than their Canon/Nikon counterparts.
Weight-wise it falls within Polarie's weight limit but its APSc so less light gathering ability and compared to my D800E is quite noticeably less bright for the same exposure. Still its a great little camera for this sort of work. Fuji also makes an XM1 (even smaller, lighter and less expensive) and an XPro 1 plus a fixed lens X100s (often compared to Leicas).
Here are the results of several exposure lengths. I used F4 mostly as there is some slight distortion in the very corners of the 14mm F2.8 lens. The Fuji 16-55 F2.8 zoom does not have this problem so I may use that next time and sacrifice a little bit of field of view.
I aligned the Polarie at first with the compass/tiltmeter accessory. I noticed some star trailing at 10 minutes so I got the polar alignment scope and did it more accurately. The 15minute shot shows very little star elongation.
http://www.pbase.com/gregbradley/image/152153598 15 minutes with 2hours of CCD Ha added ISO 200 F4
http://www.pbase.com/gregbradley/image/152153619 30 seconds ISO 3200 F4
http://www.pbase.com/gregbradley/image/152153619 10 minutes ISO800 F4 Ha areas starting to show subtle colour
http://www.pbase.com/gregbradley/image/152153765 210 seconds ISO1260 F4
http://upload.pbase.com/image/152167635 45 seconds ISO6400 F3.2
You can notice much deeper colour saturation and the Ha areas starting to show at 10 minutes. The Milky Way is bright even at 30 seconds F4 and ISO3200 but colour is a bit washed out and lacking in the Rho area.
Perhaps not so noticeable is the lower amount of noise in the longer exposures despite no noise reduction whatsoever.
Greg
