30
\$\begingroup\$

One of my longstanding interests is long exposure night photography.

However, many of the situations where I am shooting require exposure lengths on the order of 10 minutes, which is far beyond what the internal meter in my camera (5DII) can handle.

Right now, I take a test exposure, extrapolate from that to guess my exposure time, try again, etc...

This works, but since the test exposure alone can be 10 minutes, and it generally takes me a few shots to get the exposure right, it can take 30 minutes to an hour just to figure out exactly how long I need to expose for.

Obviously, this is really inconvenient (fortunately, long exposure subjects don't move much, most of the time).

For instance, recently I did a series of 13 minute exposures at f4, ISO125. It came out really well, but I would up spending 2 hours to get one shot.

Anyways, is there an easier way to determine how long an exposure has to be, preferably that doesn't take multiple test exposures? I've looked at light meters, but none of them make explicit statements about how low a level of light they can handle.

\$\endgroup\$
6
  • 4
    \$\begingroup\$ Is the problem that the meter simply refuses to report long exposures, or that it meters wrong? In the former case, have you tried to meter with a very high ISO? E.g., if you want to shoot with ISO 100, then you could meter with ISO 3200 and multiply the time by 32? \$\endgroup\$ May 15, 2011 at 10:13
  • 2
    \$\begingroup\$ @Jukka Suomela - The meter on Canon DSLRs goes up to a max of 30 Seconds. Anything longer then that, and you HAVE to use bulb mode. I'm timing my exposures with a stop watch and a remote release. \$\endgroup\$
    – Fake Name
    May 15, 2011 at 10:24
  • 4
    \$\begingroup\$ So did you try the trick of metering with a very high ISO? For example, 10 minutes with ISO 125 ≈ 23 seconds with ISO 3200, so the meter should be able to report the time (whether the result is correct is another question). \$\endgroup\$ May 15, 2011 at 10:40
  • 4
    \$\begingroup\$ A related trick: You could do your test exposures with a high ISO. That way you can do your test shots much faster, and then you only need to switch to low ISO for your final shot. \$\endgroup\$ May 15, 2011 at 10:52
  • 2
    \$\begingroup\$ You might want to avoid ISO125 in favor of ISO160 on a 5dmkII - see the noise test by Peter Troeger - vimeo.com/20239453 \$\endgroup\$
    – Imre
    May 15, 2011 at 12:27

3 Answers 3

29
\$\begingroup\$

At light levels this low, you'll be much better off by taking some test shots and checking their histograms rather than just relying on a light meter (which is usually optimized for measuring light, not darkness). However, you can make the test shots take less time.

Perform the test shots at the maximum ISO your camera can handle (avoid the uncalibrated expanded ISOs though), and multiply the measured proper shutter time by the factor you decrease ISO for your real shots.

Learn how a histogram's right-hand end for shots underexposed by 1/3 steps, 2/3 steps, 1 step looks like, so you'll recognize them and need fewer test shots.

Take the test shots with widest aperture and multiply measured shutter time by the difference to real aperture squared (another way to say it: multiply by two for each stop you'll be closing the aperture).

For example, if your test shot showed that at f/2.8 and ISO 6400 your exposure should be 8 seconds, then at ISO 125 f/4 you will need to expose for 8 * (6400 / 125) * (4 / 2.8) ^ 2 = 8 * 51.2 * 2 = 819 seconds = 13 minutes 39 seconds.

Note that if you happen to be shooting film, you'll also have to adjust for the reciprocity failure of the film you are using.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ The expanded ISOs are not "uncalibrated" as you claim. They are merely digitally boosting/dividing the values from the sensor readout, instead of controlling the analog amplifier. \$\endgroup\$
    – Nayuki
    Feb 6, 2017 at 0:13
5
\$\begingroup\$

If you've got an Android phone then there's an Exposure Calculator app available. You can download it from the link and it's also in the Android store (it's free).

\$\endgroup\$
-1
\$\begingroup\$

I don't think there exist any technology that can give you exact metering for very long exposures. Most people who shoots long exposures do this on the trial-and-error basis, till you get the perfectly exposed picture.

Metering itself is very tricky and people always shoot photos that the camera marks as either underexposed or overexposed, but you know its perfectly exposed when you look at the picture. So, even having a technology able to meter very long exposures wont be that much helpful I believe.

I also believe, as you keep spending more time finding the right exposure, your efficiency will increase and you'll be able to settle on the exposure settings quicker. Learn from your mistakes and make your exposure guesses closer to perfect by practicing more.

\$\endgroup\$
3
  • \$\begingroup\$ I'm definitely getting better at guessing exposure times, but it is still time consuming. It just seems there should be a better way, I don't know, a photomultiplier based sensor or something. There are light sensors that are very sensitive. \$\endgroup\$
    – Fake Name
    May 15, 2011 at 10:25
  • 5
    \$\begingroup\$ @Fake Name: This answer is terrible advice, and there is a better way! I shoot very long exposures for astrophotography all the time and I am able to predict my exposure times with excellent accuracy by using a little but of math in a procedure similar to what @Imre outlined... \$\endgroup\$ May 15, 2011 at 15:22
  • \$\begingroup\$ Yes, Imre's answer is good! I was always bad in math :( \$\endgroup\$ May 16, 2011 at 4:20

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.