10-02-2013 11:52 AM
The best way, is if the moon is out, to focus on that. If the moon isn’t out, and you have no experience with the lens, is to set the focus somewhere between the sideways “L” and the infinity mark. I have no experience with the 24L, but I’m sure someone can weigh in. Better yet though, is to take your lens out during the daytime (or dusk) and focus on the horizon. Have a look at the meter on your camera and see where it is. Then just set to there when you’re shooting stars.
Keep in mind that the earth is moving though. Don’t mistake motion blur for OOF.
10-02-2013 12:43 PM
Thanks for the tips. Yes I know the earth is moving. With this wide angle lens the focus point are so small. I'm going to use a 90 degree viewfinder attachment that magnifies. Also I can zoom in on live view during the shoot.
10-02-2013 12:49 PM - edited 10-02-2013 12:49 PM
Live view probably isn't going to work for that. I recommend just taking it out in the day time and looking at where it focuses to infinity. That's why that little meter is there, just line it up manually when you're out shooting and it'll be in focus every time. It'll either be on the infinity mark, or just before it. If you can move your camera past the infinity mark then it will probably be on it. Canon puts some extra room at the end so auto-focus doesn't slam up against it when focusing at infinity.
My comment on the earth moving wasn't meant to be snooty. You realize that if your shutter time is too long that you will see movement, correct? And that movement, when shooting stars, looks quite like being out of focus. People new to astrophotography often don’t realize this, and times as “short” as 10 seconds can produce unsharp stars. At 24mm you’re probably going to be able to get away with twice that, but still, my point stands.
10-02-2013 12:58 PM
Manually dial the lens to the infinity focus point. Note this looks like a sideways "L" and the actual focus point is the intersection at the base of the "L".
The lens will turn beyond that point -- so don't just turn it until it stops. Different wavelengths of light have a different focusing distance so the lenses are designed to focus beyond infinity to compensate for this. If the lens were used on a film camera shooting infrared film, then the far right side of the "L" becomes the focus point -- but not for visible light.
When I use a telescope, we use something called a Bahtinov focusing main to obtain perfect focus. But this requires a magnified bright star and I suspect it wouldn't work for a wide angle lens. The mask works by deliberately creating diffraction spikes on a star. Two spikes are in the form of an "X" and a third is in the form of a vertical bar "|". As you focus, the bar slides left/right across the X. When the bar is exactly centered in the X you have achieved perfect focus. You can find YouTube videos demonstrating the effect and there's even a site that generates a mask template (you can print it on your printer, cut it out and place it over a lens to focus.) Again... I suspect it probably wont work on a 24mm lens because the angle is so wide that you may not have a single star in the frame bright enough. The brighter the star, the longer the diffraction spikes.
One of the nice things about focusing on anything in space is that once you focus on any object, every object is in focus. E.g. if the moon is in tack-sharp focus, then so are the stars. When focusing our telescopes, we point to a bright star, use the mask, get the focus perfect, lock the focus down so it can't move, then point the scope back to the faint fuzzy object we want to image.
If the exposure of the stars runs too long then the stars will elongate because the Earths rotation causes them to appear to be moving. There's a rule for this, sometimes called the "rule of 600".
For a full-frame camera you can generally use 600 ÷ focal length = exposure time limit
If you have a full-frame camera, you can use 600 ÷ 24 = 25
That means if you run exposures longer than 25 seconds you'll start to notice your stars are no longer rounded or pinpoint, but are beginning to elongate and leave a trail.
If you have an APS-C DSLR the image sensor is smaller and much of the image projected into the sensor body isn't used. It's as though you are using a longer focal length lens than you really are. That means the formula has to get adjusted.
If you divide 600 by the sensor crop factor (which is 1.6 for all Canon APS-C bodies) you get 375 (if it's easier to round off to 400, go ahead... you will not likely see much difference.) That would give you 375 ÷ 24 = just under 16 seconds (16 would be fine).
The rule of 600 is a good baseline. I have cheated this up upon occasion and while I get away with a tiny bit of cheating, if I cheat too much I notice.
Of course if it's your intention to have star trails... you can ignore the rule.
10-02-2013 03:57 PM
10-02-2013 05:56 PM
great information. I'm shoting with a 5D lll. The wide angle is the issue, I'm actually shoting northern lights while trying to make the stars in super focus. A chalange for sure. I have gotten great shots in the past but they were out last night and could not get it done. I'm adding a 90 degree viewfinder with magnification and also try zooming in while in live shot mode.
10-02-2013 07:41 PM
The guys in our astronomy club use Rebels with the articulated back for that reason. Being hard to see through the view finder and the Live View works well.
A 24mm lens is good for wide-view sky shots. But if your Northern Lights are very much brighter than your stars, you are correct. It will be a challenge.
Have you ever heard of "stacking"?
10-02-2013 07:53 PM