I am trying to get into long exposure time lapse photography, and was wondering if there was a way of speeding up or turning off the processing of the image! When doing long exposure time lapse at (for example) 15 second exposure, once the image has been taken the camera then processes the image which can take a further 15 to 20 seconds before it allows another photo to be taken! Can this processing time be sped up or turned off to allow shooting to be quicker?
2 Answers
To reduce the processing time for long exposures, you want to turn off Long Exposure Noise Reduction. However, you may not want to give up the benefit of LENR.
Long Exposure Noise Reduction (LENR) is Canon's nomenclature for in-camera dark frame subtraction. When you take a photo the camera will expose the image normally and then use the same settings to create a dark frame with the shutter left closed. The readings for each pixel in the dark frame will be subtracted from the reading for each pixel in the first frame before sending the raw data to your memory card.
The time required for a dark frame is the same as the time required for the initial exposure - so if you shoot a 30 second exposure you will then have to wait an additional 30 seconds before you can take another shot.
If you disable LENR then you'll find that long exposures will very likely demonstrate more pattern noise (read noise) and hot pixels. You can manually take a dark frame at the end of your session (just shoot a frame at the same settings with the lens cap on and the viewfinder blocked from any stray light). You can then use post processing software to apply the dark frame to your exposed frames.
-
1\$\begingroup\$ Thanks for your advice, you have been more than helpful 👍 \$\endgroup\$ Commented Apr 30, 2017 at 12:03
-
1\$\begingroup\$ Would it help to capture two dark frames, before and after the shooting session, and to use an average of the two? Or will the pattern noise be expected to be constant over this timescale? \$\endgroup\$ Commented May 1, 2017 at 10:15
-
\$\begingroup\$ As the sensor temperature increases over the course of the session, pattern noise will increase. If the session is long enough, then dark frames should be made at periodic intervals and applied to the frames preceding each sample. A lot of more detailed information is contained in many of the questions/answers that result in searching this site for the term "dark frame subtraction." If one wants to use multiple samples of dark frames to eliminate the random portion of the result, a discrete set should be taken at each stage. \$\endgroup\$ Commented May 1, 2017 at 14:21
-
\$\begingroup\$ Die you mean "turn on" or "turn off"? (1st sentence). \$\endgroup\$– ZenitCommented May 3, 2017 at 10:27
It's not actually processing for most of that extra time. It is taking a second exposure with the shutter closed, for dark frame subtraction. This removes sensor-based pattern noise. Of course, there is a bit of processing involved in the subtraction itself, but most of the time is in taking a second exposure with the same shutter time as your actual one, to make the noise pattern as similar as possible.
In your camera's settings, find "Long Exp. Noise Reduction". You can set this to off if you like. That can be useful if you are taking a series of photos with similar settings (as you probably are for your time lapse). You can take the dark frame yourself manually — lots of post-processing software can use one. Or, you can just accept the probability-greatly-increased noise.
-
\$\begingroup\$ Is this true only of jpegs, or does the same thing happen for raw images? \$\endgroup\$– NoahLCommented Apr 30, 2017 at 19:55
-
\$\begingroup\$ Yes: photo.stackexchange.com/questions/2691/… \$\endgroup\$– mattdmCommented Apr 30, 2017 at 23:13
-
\$\begingroup\$ @NoahL From the other answer to this question (last sentence of the second paragraph): "The readings for each pixel in the dark frame will be subtracted from the reading for each pixel in the first frame before sending the raw data to your memory card." \$\endgroup\$ Commented May 1, 2017 at 0:19