Help with star trails

invisible · May 22, 2008

Tomorrow I'll go out to the lake to try and shoot some star trails. It's going to be my first time doing this kind of long exposures, so I'm hoping some of you can help me out with tips.

My camera is a Nikon D70s with the Nikkor 18-200 lens. I'm bringing my tripod and my cable release.

- How many hours of long-exposure shooting will I get with a fully charged battery (EN-EL3a, Li-ion)?
- What's the longest exposure that the D70s can stand? (I have read about the dangers of overheating the sensor, but I couldn't find data specific to my camera.)
- I'll probably go the widest possible angle (18mm), to catch some of the surroundings, and definitely go with the minimum possible ISO (200). What aperture/speed do you guys recommend? White balance?
- What about mirror lockup? Will it make a difference?

Any other tips will also be appreciated.

Thanks so much for your time!

Garbz · May 23, 2008

Don't even bother. An exposure longer than about 15 minutes will be so full of noise and dead pixels that it will be almost unusable.

But that doesn't stop star trails. Just change the technique. Setup your camera aperture wide open or close to it and ISO200, and take a constant series of 2 minute exposures. Make sure long exposure NR is off, so there is no gap in between the photos you take. Then use an image stacking program http://www.tawbaware.com/imgstack.htm to generate a final composite. You will actually get (shock) a usable image. But note your sensor still heats up. Maybe you want to take 1 minute on and 10 second gaps between your exposures. If you zoom in this causes gaps in the star trails but at a wide angle 10seconds won't give you a noticeable gap for any normal sizes (reads as not zoomed at 100%)

I expect the D70 should get the same time than the D200 so you can get about an hours continuous exposure. 2 with a battery grip. If you split up and add a 10 second gap every 60 seconds of course your total shooting time goes up along with your trail length.

Ignore mirror lockup. It makes sense only for shutter speeds between 1/30th and 1/2 second. Beyond that the vibration is so small in the final exposure it is not visible.

Finally I hate to say it to the digital buffs here, but if you really want to take star trails, film camera, bulb cable release. You can go all night without even the slightest concerns for noise or camera damage.

invisible · May 23, 2008

Hey Garbz, thanks for your great input!

Even though I think I'll stick to traditional long exposure tonight, would you be so kind as to post a couple of pictures of startrails processed with the method you described? Also, if you have some, can you post pictures shot the traditional way, with the noise and dead pixels you said I should expect?

Thanks again!

Garbz · May 23, 2008

Certainly:

This image was processed from 200 30 second frames. With a 5 second gap between frames. Now the lack of stars come from my light pollution floor. These images were stacked with a MAX algorithm where each image is analysed and the brightest luminance value is taken for the final image. This has the advantage of trailing the stars without increasing the light pollution. Using a standard ADD function to get the equivalent of a 100minute exposure would have resulted in an image probably so blown out it would have been unusable. Even though there light trails show 2 hours of movement, the light pollution I think in the final image came from a 2 minute exposure. (30 seconds made the background unnaturally dark compared to the stars).

I can't post an example of the other one because the image I got out after a 1 hour exposure was so noisy that after an hour in photoshop with the clone too and trying various filters I totally gave up on it and just deleted it. Take a 15 minute exposure and count the dead pixels, and then remember the number of dead pixels increase logarithmically in time (or close to it based on observation). I am afraid to say but the noise in that final picture I got could probably be measured as a percentage of dead pixels it was that bad.

Cinka · May 23, 2008

Garbz said:
Certainly:

This image was processed from 200 30 second frames. With a 5 second gap between frames. Now the lack of stars come from my light pollution floor. These images were stacked with a MAX algorithm where each image is analysed and the brightest luminance value is taken for the final image. This has the advantage of trailing the stars without increasing the light pollution. Using a standard ADD function to get the equivalent of a 100minute exposure would have resulted in an image probably so blown out it would have been unusable. Even though there light trails show 2 hours of movement, the light pollution I think in the final image came from a 2 minute exposure. (30 seconds made the background unnaturally dark compared to the stars).

I can't post an example of the other one because the image I got out after a 1 hour exposure was so noisy that after an hour in photoshop with the clone too and trying various filters I totally gave up on it and just deleted it. Take a 15 minute exposure and count the dead pixels, and then remember the number of dead pixels increase logarithmically in time (or close to it based on observation). I am afraid to say but the noise in that final picture I got could probably be measured as a percentage of dead pixels it was that bad.

Wow! That's pretty amazing.

manaheim · May 23, 2008

Awesome pic, Garb.

FYI, the D200 is somewhat notorious for being "not so great" on battery life. You may find the D70 will shoot for quite a while. I would spend a whole evening (4+ hours) shooting with my D100 and have no problem at all. (granted, I was moving from subject to subject)

Garbz · May 23, 2008

Continuous exposures are a very different beast. I wouldn't say the D200 is bad by any stretch. With a battery grip I have used my camera for days without actually recharging.

The 3 components that chew the most power are the light meter, the LCD, and the sensor in order from least to most. When shooting normally most of these are on for no more than 10 seconds at a time before the camera shuts them off and puts itself in idle mode to save power.

I guarantee your D100 would not 4 hours continuously exposing. Sensors chew a huge amount of juice. Which is why P&S cameras and cameras with live view turned on have a despicable battery life.

Socrates · May 23, 2008

invisible said:
My camera is a Nikon D70s with the Nikkor 18-200 lens. I'm bringing my tripod and my cable release.

You'll have a heck of a time attaching a cable release to the D70s. There's no place to connect it! Instead, you should use the ML-L3 wireless release.

Socrates · May 23, 2008

Garbz said:
Ignore mirror lockup. It makes sense only for shutter speeds between 1/30th and 1/2 second. Beyond that the vibration is so small in the final exposure it is not visible.

He'll have to ignore mirror lockup. The D70s, like my D80, supports mirror lockup only for filter cleaning. You can't use it to take pictures.

K_Pugh · May 24, 2008

As far as i remember the D70 can only take an exposure of 30 minutes on battery power? I sort of recall something like that from my manual a few years ago - not entirely sure.

invisible · May 24, 2008

Garbz, thanks again for your comments. I'm back from shooting and everything you said was verified in the field. In total I took three long exposures: one 30-minute shot and two 10-minute shot. The 30-minute shot is garbage (even with noise reduction). The other two are fine. Maybe next time I'll try your method.

K_Pugh said:
As far as i remember the D70 can only take an exposure of 30 minutes on battery power? I sort of recall something like that from my manual a few years ago - not entirely sure.

Well, the long exposures I took tonight total 50 minutes. Prior to that I shot some sunsets, maybe 90 photos. The battery indicator still shows as full.

Socrates said:
You'll have a heck of a time attaching a cable release to the D70s. There's no place to connect it!

Actually there is. I have an MC-DC1 cable release, and I plug it to the camera every time.

PhotoDonkey · May 24, 2008

I think this would be the kind of thing that I'd bust out my film SLR for. Digital shooting is great, but there are still some things that film is going to be naturallly better suited for.

Socrates · May 24, 2008

invisible said:
Actually there is. I have an MC-DC1 cable release, and I plug it to the camera every time.

Oh yeah, I see it now.

manaheim · May 25, 2008

Garbz said:
Continuous exposures are a very different beast. I wouldn't say the D200 is bad by any stretch. With a battery grip I have used my camera for days without actually recharging.

The 3 components that chew the most power are the light meter, the LCD, and the sensor in order from least to most. When shooting normally most of these are on for no more than 10 seconds at a time before the camera shuts them off and puts itself in idle mode to save power.

I guarantee your D100 would not 4 hours continuously exposing. Sensors chew a huge amount of juice. Which is why P&S cameras and cameras with live view turned on have a despicable battery life.

I don't think any of our cameras would do 4 hours continuously. (I mentioned in the 4 hour thing that I was moving around a lot)

Anyway, my battery comments come from a combination of things. One anecdotal, one not.

Anecdotal: I had a D100 for about 5 years and shot with it constantly so I got used to a certain amount of battery life. When starting some of my pro work I borrowed by buddy's D200... I was quickly surprised as to how fast the battery wore out.

Not: (I can't believe I'm going to post a Ken Rockwell link)

http://www.kenrockwell.com/nikon/d300.htm
http://nikonglass.blogspot.com/2008/01/d300-impressions-from-d200-user-pt4.html

There I got one more in there to make up for the Ken link.

I had also read somewhere that Nikon just hadn't gotten the internals of the D200 down well enough and it was just naturally a bit inefficient on power... they fixed it in the D300... but I can't find that link.

Not that any of this necessarily matters that much... you have whatever cam you have and become aware of how it performs... it's just interesting.