Do you use Long Exposure Noise Reduction?

[millan]

When shooting with ND1000 or infrared filters, I have dilemma whether to set the LENR on, or off. Off means faster work, on – less noise. What do you prefer? BTW, is also RAW file less noisy when LENR is applied?

 

[Loswr]

Long exposure NR is one of the few settings that actually affects the RAW file.

The visibility of image noise in long exposures is dependent on the scene. Shots with a 10-stop ND filter in daytime usually yield a bright scene where noise isn't evident, meaning LENR isn't needed.

Try both and see what works for you...

 

[privatebydesign]

When shooting with ND1000 or infrared filters, I have dilemma whether to set the LENR on, or off. Off means faster work, on – less noise. What do you prefer? BTW, is also RAW file less noisy when LENR is applied?

I don't use filters, but do use LENR when taking longer exposures, particularly when the image is dark, or low key, as colour noise becomes an issue.

Having said that you only need one dark exposure and then just subtract it from multiple images of the exposure time is the same. So if you wanted to do a series of images to make a star trail, for instance, take a set of 30 second exposures and one 30 second dark exposure and then just subtract that one from all the others.

 

[wayno]

No. Never. I have never found much use for it and much of my work is long night exposures.

 

[Don Haines]

I tested my 60D with and without LENR and found that for dark skies there was less noise with LENR off.

 

[Pi]

I tested my 60D with and without LENR and found that for dark skies there was less noise with LENR off.

Right. LENR subtracts a black frame from the original, in an attempt to cancel hot pixels. The effect is that all "noise" which is repeatable (and in that sense, not random), disappears. Random noise however (random in time, not in space) gets amplified by sqrt{2}. The subtraction cannot be perfect BTW - it would produce more negative values than the RAW format can handle and the camera probably rounds them up.

I turn LENR off as well, and handle hot pixels manually. Some converters would mask them automatically anyway.

 

[dgatwood]

I tested my 60D with and without LENR and found that for dark skies there was less noise with LENR off.

Right. LENR subtracts a black frame from the original, in an attempt to cancel hot pixels. The effect is that all "noise" which is repeatable (and in that sense, not random), disappears. Random noise however (random in time, not in space) gets amplified by sqrt{2}. The subtraction cannot be perfect BTW - it would produce more negative values than the RAW format can handle and the camera probably rounds them up.

I turn LENR off as well, and handle hot pixels manually. Some converters would mask them automatically anyway.

Seems like it would be better if the camera contained a series of profiles indexed by sensor temperature, with each profile generated from the average of a few thousand consecutive 30-second black shots. Averaged over a large enough number of black shots, random noise should converge towards zero and static noise should converge towards a particular value.

 

[wayno]

I used to get hot pixels from jpg shooting but have never seen one when shooting in RAW in LR. I've assumed LR sorts it out on conversion. I refuse to believe my 5d2 is impervious to it but I've never seen one in nearly two years of relentless night shooting.

 

[Pi]

Seems like it would be better if the camera contained a series of profiles indexed by sensor temperature, with each profile generated from the average of a few thousand consecutive 30-second black shots. Averaged over a large enough number of black shots, random noise should converge towards zero and static noise should converge towards a particular value.

I have noticed weird things like hot pixels one night, in a series of shots; and lack of them or hot pixels somewhere else later the same night or the next evening.