Lighthouse

[jkshyt]

_A748404-2-Enhanced-NR by jk Sullivan, on Flickr

[Round Tuit]

I'm sorry but this does not work for me. I think that you would need a bit of fog to reveal the beam of light or else a starry night to break up the large expanse of featureless darkness.

[royent]

Too much noise particularly in the sky
Roy

[Geoff F]

Because of the noise level in a rather drab sky I would consider cropping away about a quarter to a third of the sky with also some side cropping. Many options could be considered.

With high noise levels in some of my difficult images I usually do a substantial noise reduction on a duplicate later then merge it with the original image so the main subject remains as originally shot but the problem areas get cleaned up.

[DanK]

I think the level of noise in this photo makes it unsalvageable. I don't think any noise reduction program can manage this, but I may be wrong. I don't think I have every seen the square noise patterns that show up here, and I don't know what might cause that, as opposed to random noise.

In my experience, night photographers shooting static subjects usually use low ISOs and long exposure times. You can get very long exposures, although smaller sensors will eventually overheat and generate horrible noise. I'll post below an image shot at ISO 100 for 7 minutes, using a Canon 5D Mark III, just to illustrate how little noise results. I didn't use any noise reduction.

The one problem here is that if you want to get the light shining in one direction, you need a short exposure. I think the only way to do that would be to blend multiple images.

[jkshyt]

Terrible image, I shot it handheld at a very high ISO and used LR auto noise reduction.
I'll have to redo this one more time.
Any ideas on how to capture a lighthouse?

[Geoff F]

As Dan has already mentioned you need a lower Iso setting. I don't like going over 800 or maybe 1200 in some circumstances.

This means a longer exposure with a reasonably open aperture. What lens do you have? Is the light fixed or a flashing beam which requires close timing?

A tripod might help but your shutter speed will depend on the light beam.

Clear air, not hazy will be an advantage.

Your general composition, with only a narrow dark band at the bottom, is good. Experiment and take several shots with different setting.

[escape]

Hi Jack. Will you be able to reshoot the lighthouse? The image is not terrible, in my view; just the first step in a series to get the image you're after. Hoping you can reshoot.

[DanK]

I'd say that a tripod is essential at that light level.

High ISO is the setting that caused the noise, but the real issue is exposure. When you boost ISO to a very high level, you are exposing the sensor very little and then amplifying the result. Because noise is nearly constant, this results in a very poor signal-to-noise ratio. ISO is also relevant for night photography because as you raise ISO, you reduce dynamic range.

[billtils]

as you raise ISO, you reduce dynamic range.

And you have to deal with the bright beam of the light and the darkness everywhere else. This is NOT an easy shot to pull off, but consider using a tripod (if you have one) and bracketing your exposures, perhaps starting with 3 stops under to 3 stops over.

[DanK]

What makes this image complicated is the rotating light. If you want to maintain the directionality, you can't have an exposure longer than the light is on that side. That exposure is probably a tiny fraction of what you would need for a clean image.

Without having the actual values, it's impossible to know, but you might be able to get a much better result while maintaining the directionality by using photo stacking rather than a long exposure. You can find an explanation here: https://digital-photography-school.c...hoto-stacking/. The idea is that you take a bunch of identical images with the longest exposure that retains the directionality. You would then use one of several stacking methods to average those images. The logic is that the noise is random, while the signal isn't. If you average enough images, the noise will cancel out. However, in this case, the underexposure may be so severe that this doesn't work well without a prohibitively large number of images.