Get rid of noise in dig' time exposures

[mikephotog]

Interesting technique

http://www.dpreview.com/learn/?/Image_T ... ots_01.htm

[losfp]

So basically taking a "blank" frame and using that as a template to eliminate sensor noise in your actual image. I really should try that one day, but I don't do many time lapse exposures, so I normally forget to take the second shot!

Though I think some cameras have an option to do it at the point of taking the shot (I think the D70 does it? Not sure, will check when I get it).

[Alpha_7]

The D70 does have this options, you can turn it on in the menu. Bear in mind it takes twice as long to take a shot.. as you have a 20sec exposure. then 20 sec blank exposure... so for some applications this isn't ideal.

[marcotrov]

I guess Graig you'd only consider it sparingly and where you would be in a situation, apart from the long exposure, where you would expect high probability of noisy outcome. Where is this option on the D70 craig?
cheers
marco

[Manta]

Marco,
You'll find "Long Exp. NR" in your setup menu and can set it to on or off. I've used it before for star trail shots - 30 minute exposure plus 30 minute noise reduction. Make sure your batteries are fresh!!!

[marcotrov]

Thanks Simon
cheers
marco

[Gordon]

Interesting that he calls it a "Flaw Frame" for the past ~20 years they have always been known as dark frames by astronomers.
Astronomers use long exposures, and for CCDs that arent cooled to -80 or -100C there is a thermal signal that needs to be removed. An exposure of equal length to the light frame at equal temperature with no light falling on the chip is what the dark frame is. There is also the bias frame, this is obtained with an exposure of zero length, just reading out the CCD. That signal also has to be removed. In digital cameras the 2 are combined in the noise reduction exposure.
This does not compensate for the variations in sensitivity of the pixels across the chip, or dust shadows, which astronomers remove by producing a frame to compensate for the variations, called a flat field. This can be obtained by taking images of the twilight sky, or a white board evenly illuminated, or even by adding together dozens or hundreds of images of the night sky, the data images, and then producing an average of them. It is then divided by a number to produce an overall level near 1. The light frames, which have the dark and bias removed from them are then divided by the flat field frame, which produces the calibrated final image.

The noise reduction frames in digital cameras are just bias+dark frames. The "dust off" frames are flat fields. The only problem with them is they have to be taken at the same f/# as the light frame, because the dust shadows vary with aperture.
However, there is another problem- the bands of noise that appear on frames with low overall levels, and at higher ISO settings, especially when the camera is warm.
These seem to be in random places and cannot be removed by any of the methods above.
They are in fact, a real PITA!

If someone figures out how to remove them, please let me know!

Gordon