I'm going up to the Highlands of Scotland (Loch Ness) next week for a week long break. This is really a break for my wife and I, she is 7 months pregnant with our first and this is likely to be our last holiday alone together until we retire!

There isn't much daytime and I don't want to make this an exclusively photo trip, but given just how much night there is together with the clear air and sky there I wanted to invest time in some star trail photography.

I intend to use my Elan II, DSLR's and 3 hour exposures arn't the greatest of combinations. I don't have a timer for it which is a shame (I'm consulting with an engineer who is interested in making a cable release with two seperate buttons for AF and shutter release, a timer, an adaptor to use it with either canon connection and a plug in Inverlometer, should be good, anyone interested in buying one for less than the canon one?).

Couple of questions about star trail photography. Everywhere I look they say to use a fast lens wide open. Would it not be better for sharper pictures to close down and keep the shutter open for longer? With 10 or so hours of pitch dark this shouldn't be a problem.

I have a 24-70L lens, I also have access to a 24mm f2.8, which would be better for this type of photography at that focal length given that the prime has less elements?

What kind of battery life can I expect from either 4 AA's or the CR25? Will it be enough for 3 hour exposures in the cold?

Some people suggest using Neg film for more latitude and compressed contrast, some say that only slide film will do, any opinions?

I've pushed my 20D up to two hours so far without any serious problems.



My 20D with the grip and two BP511 batteries seems to be good for four or five hours of exposure time.



These two guys were film shooters...

www.lostamerica.com
www.nightphotographer.com

...now they're 20D owners.

Here's an article that may help a bit:

http://www.digiteyesed.com/clippings/f2_lp_article.pdf

Hello Pom,

It has been quite a few years since I visited Scotland, but I have very fond memories of several trips to the far north, including one right at this time of year.

I happened to watch a video recently on night photography that had a worthwhile suggestion for digital photographers--before you make a long exposure, try shooting one at the camera's highest ISO setting without noise reduction shot wide open. Even in a very dark location, you probably will get something in less than 30 seconds, the usual maximum for the camera's automatic metering. The resulting image will certainly be too noisy to be useful ( and obviously it won't show star trails) but it will reveal where the light sources are (even those that aren't apparent to the eye) and it will give you a useful starting point for estimating the long exposure at low ISO. Fortunately digital shooters don't have to be concerned with reciprocity. It sounded like a useful tip to me.

I hope you have a great trip.

Hi Pom

I tend to agree with you - DSLR's are fine for night photography with exposures in the minutes range, or perhaps Digiteyesed's couple of hours, but for long star trail photography I'd stick with film.

This is probably my most successful effort.

Click to view attachment

I was looking for star trails, but also wanted some foreground interest, so I chose a night where the moon would rise around 11pm. I used an exposure of approx 4.5 hours at f7.3 and the moon rose 20 minutes or so before I closed the shutter. This gave the light on the foreground. Film was Provia 100F. Lens was 16-35L at 16mm.

Fortunately my EOS3 does not require battery power to keep the shutter open, but I'd think you should be able to get a decent exposure time out of a battery, even in the cold.

I'd think about star trail exposure like this. Exposure time controls the 'length' of the trails you get and aperture controls the brightness of the trails. Exposure time (shutter speed) won't affect the brightness of the trails. But stop down too much and you'll only get the brightest stars, open up too much and you'll get a mass of them - probably too many. For 100 ISO I'd reckon on f5.6 to f8. In the end it's all trial and error...

I've got the Canon intervalometer, but a standard release is fine. You've just got to set the alarm, wake up and go and unlock it!!

Hope this helps a little. All the best for the break and for the arrival in a couple of months.

Regards

Frank

I didn't know that canon made an intervalometer for it's modern DSLR's, can you point it out to me?

In terms of noise reduction wouldn't it be better to use a series of shorter exposures, then blend them in Photoshop, or are the new sensors on the Canon 20D and 5D so good that this is not a concern?

The 20D's predecessor even at iso100 seems useless, because of noise, at any exposure over 60 seconds. The camera I am refering to has about 10000 exposures on it. Perhaps there is a possibility that sensors get worse (only in terms of long exposures) with usage?

One thing I have been tempted to try (but have not yet done), is to take a black level exposure (of the same shutter time) with the lense cap on before taking a long night exposure. Then subtracting this black level exposure from the actual exposure. If the noise is random then this will do nothing. However if the noise is correlated between frames, this should remove a large amount of it.

- Timothy Farrar : farrarfocus.com

Pom

It's the TC-80N3. Michael reviewed it here a few years ago.

It connects via the later 3 pin plug which I don't think the EOS 50/Elan II has, but I seem to recall that an adaptor is available.

I don't doubt that it's possible to construct a star trail shot from multiple digital exposures, but I can't see the point. For me it's about the only area that film retains an advantage.

Pom,

Take a look at this tutorial recently posted on the Radiant Vista web site:

http://www.radiantvista.com/tutorials/video.php

It is a pretty good intro to night photography and talks about the "test" shot technique referenced in an earlier post by jdemott.

Have fun and good luck with parenthood!

I am now a grandfather - it's all worth it.

David

My Photo Web Site

FWIW, here is a shot I took recently. 3 hour exposure, EOS 1V, Velvia 100F slide film, EF 24 f/1.4 lens, f/8.

My 300D with over 20,000 exposures on it has not aged noticeably in terms of sensor noise.



If you like reciprocity shift/failure. There's nothing stopping you from taking four or five one hour photos with a Canon or equivalent and stacking them in PS. There's no way I'll ever consider going back to film for night photography now that I've tried it with digital.

Note: The batteries in a 20D aren't good for beyond a couple of hours. I carry along a Canadian Tire Eliminator power source and an ACK-E2 AC adapter for my 20D so I don't have to worry about running out of juice.

Here's a good Dan Heller article on photographing star trails:

http://www.danheller.com/star-trails.html

Here's a program that does this for you so you don't have to mess around in photoshop. I haven't used it for startrails, but I've used it to make time lapse videos. It seems to work for that purpose. Something to look at anyway.

http://www.startrails.de/html/software.html

Bryan

A friend with a lot of experience shooting digital star trails also mentioned this as being the best method.

Besides, the key is to use a wide aperture and high enough ISO value (400 ISO being a good compromise).

I have tried a little bit with 4*5 Ebony using a Schneider 110 opening at 5.6 with Provia 100F pushed 2 stops.

Regards,
Bernard

I haven't seen the performance of the 5D firsthand, but I have no reason to doubt that it is any worse than my 20D, which is superlative for long exposure star trail photography (exposures upwards of an hour at ISO 200).

There are plenty of things I still love to use film for, but you'd have to put a gun to my head to make me shoot star trails using film now that I've experienced digital.

It would be better for a sharper picture to stop down. However that isn't the problem. The stars move across the sky, so they aren't really in any one place for very long to register on your sensor/ film.

I've made the mistake of shooting at f 16 and higher - very little in the way of visible trails. The length of the exposure doesn't change that much, because you care about the amount of light per period that the star is over the particular pixel/ bit of film, not the total exposure time.

Total exposure ~= length of trails
Aperture ~= brightness of each part of the trail

I've found that ISO 400 at F4 gives good results on a 1DII. Just be careful to focus at infinity and not just all the way around on the focus dial. For ISO 100 film, something like f5.6 or f8 should give good results. I use the TC-80N3 and like it a lot, not just for the long timer intervals, but the multi-shot is useful for selfportraits and the like.

This is one of the interesting threads I have read for a long time.
I just run some test with my 20d and 17-40L. Even after some minutes there was massiv noise. I tried also to switch on noisereduction in RAW in the special menue.
So please let me know some exact values of your exposures and your equipment. I wanted to run some more test with my Canon 28/2,8 prime. Maybe this works better or without noise.
I also tried some test with my Singh Ray Variflex but this was really bad.

I couldn't justify spending a lot on a timer I will only use a couple times a year so I just purchased an aftermarket cable release for my 5D from an EBay seller for around $9.00.
The release button slides to lock it on. Just use the Bulb setting on your camera and use your watch for the time.

Steve.

For anyone reading this later (too late for OP), a few hints regarding noise:

What those of us who do astrophotography do is to take dark frames, the trick is to take them using same settings that you used to take the 'light' frame, for equal amount of time.

That's bad news if you are taking a 3-hour shot. So, the work-a-round is to make a Master frame collection, start building up dark frames, then averaging equal settings with equal settings, then make Master frames, keeping track of settings for later use. Just make sure that all things are the same, like temps, darkness, etc.

What I have found to be fairly good is to take anywhere from 9 to 15 dark frames, average them out and make Master frame.

To take the dark frames, just put the cover over your lens and go for it.

And yes, stacking the light frames does help...