D60 noise test

[Samir Kharusi]

I have just bought a d60, and one thing which bothers me is one
bright white (stuck) pixel on the upper left side of the frame.
This occurs at ISO 100 and above, at a speed of greater than .5
sec. on increasing ISO the intensity icreases. The position of
this pixel is constant, so I was wondering how i could do dark
frame subtraction. I have no idea on how to apply this procedure.

Any help would be very much appreciated.

This sounds like a "hot pixel". Probably the simplest way to get rid of it is simply to use a cloning tool in your photo editor at high magnification, say 300%. Since you know exactly where it occurs, just check the image you wish to work on and zap it with a cloning tool. Just pick a pixel next to it. Peferably before you have played/edited the image in any way. Extremely quick fix. Takes mere seconds.

Dark frame subtraction is for getting rid of thermal noise. Thermal noise is random in location and in timing. Ie you cannot predict where it'll be or how much there will be on any specific pixel. You take your long-exposure photo, say, one minute. Immediately put the lens cover on and take the same length exposure. When you have downloaded the two images, both with the same white balance (NOT Auto!), sharpening (Low), Saturation, etc. Simply SUBTRACT the dark frame image from the first photo. For normal digicams the noise will go down visibly. For the D60 I do not expect there will be much difference, but try it out anyway. All this must be done before you have edited anything in your images. Before sharpening, curves, etc.

Flat fielding is more elaborate and is done only when you wish to contrast stretch your image enormously, when each slight variation in the sensitivity of your sensor is magnified, thus FLAT fielding. It also gets rid of most dust problems. I have never heard of anybody using this for normal photography, even though it's pretty routine for very dim light astrophotos (neblae, galaxies, etc). Not worth chasing unless you are already very proficient in all the other steps.

 

[Skyline]

Thanks Samir for your helpful information! Guess best option is photoshop using clone then. I will try to get the camera cmos changed or mapped as the problem is evident even on ISO 100 1/60 sec !

thanks again

I have just bought a d60, and one thing which bothers me is one
bright white (stuck) pixel on the upper left side of the frame.
This occurs at ISO 100 and above, at a speed of greater than .5
sec. on increasing ISO the intensity icreases. The position of
this pixel is constant, so I was wondering how i could do dark
frame subtraction. I have no idea on how to apply this procedure.

Any help would be very much appreciated.

This sounds like a "hot pixel". Probably the simplest way to get
rid of it is simply to use a cloning tool in your photo editor at
high magnification, say 300%. Since you know exactly where it
occurs, just check the image you wish to work on and zap it with a
cloning tool. Just pick a pixel next to it. Peferably before you
have played/edited the image in any way. Extremely quick fix. Takes
mere seconds.
Dark frame subtraction is for getting rid of thermal noise. Thermal
noise is random in location and in timing. Ie you cannot predict
where it'll be or how much there will be on any specific pixel. You
take your long-exposure photo, say, one minute. Immediately put the
lens cover on and take the same length exposure. When you have
downloaded the two images, both with the same white balance (NOT
Auto!), sharpening (Low), Saturation, etc. Simply SUBTRACT the dark
frame image from the first photo. For normal digicams the noise
will go down visibly. For the D60 I do not expect there will be
much difference, but try it out anyway. All this must be done
before you have edited anything in your images. Before sharpening,
curves, etc.
Flat fielding is more elaborate and is done only when you wish to
contrast stretch your image enormously, when each slight variation
in the sensitivity of your sensor is magnified, thus FLAT fielding.
It also gets rid of most dust problems. I have never heard of
anybody using this for normal photography, even though it's pretty
routine for very dim light astrophotos (neblae, galaxies, etc). Not
worth chasing unless you are already very proficient in all the
other steps.

 

[Rudolf A. Hillebrecht]

Murphy,

That's exactly what I did, I took a look at the original. And as I
said in my post, for a four minute exposure, this is an awesome
result.

Have you tried to do a four minute exposure in the same conditions
on a D30? The differences are like night and day between D30 and
D60 results.

And I'd like to see a digital SLR that can take a four minute
exposure like this and produce a cleaner image.

jason

Hi all, together.

I have to confess that I did not read all messages of this thread, so please excuse if somebody stressed this or that anywhere else.

I part Jasons opinion. The D 60 is astonishingly noise free. After all I did not expect any digicam to come out with a long exposure result like this.

I use my D 30 in astro photography intensively (see some images of comet Ikeya-Zhang in the gallery of http://www.spaceweather.com ) and can tell you that normally one minute is the absolute limit for night shots. Some help come from addings several shorter shot images into a "long exposure", btw. astro-ccd imagers do the same.

At this point I have to add some remarks about testing: It is a complete difference to shot a test image of "nothing" (cap on i.e. simple dark image) or of a real lit object. In the second case there forms an image which mostly "covers" the noise. If you take a dark image only you simply will see the noise alone. Compare the histograms of two of such images. They are very different, so a comparison isn't truly fair that way.

Many images in astro photography are more similar to the dark image testing. There will be only few spots in it to form an image: stars. The huge rest is black (or in most cases unfortunately) artificially brightened sky. Here you will see the noise rather clearly cos you are digging in the lowest parts of your histogram.

This also is the cause, why my D 30 performs best when taking images during twilight. As long as there is enough image signal noise is secondary.

Another aspect: The tests were performed at ISO 100. This certainly would not be the choise for the astrophotographer. He will look for the highest sensitivity and use the fastest lens (I do with the 1,8/200 myself) to penetrate the skies as deep as possible. But as we know noise really can get worse at higher ISO. How abot some serious astro testing there with night sky images at ISO 800 and 1000?
I would like to get hands on a D 60 but won't have the chance so soon.

So could someone please do this for me and give us some more clues about the real astro abilities of the D 60?
TIA
Rudolf

 

[Samir Kharusi]

I use my D 30 in astro photography intensively (see some images of
comet Ikeya-Zhang in the gallery of http://www.spaceweather.com ) and can
tell you that normally one minute is the absolute limit for night
shots. Some help come from addings several shorter shot images into
a "long exposure", btw. astro-ccd imagers do the same.
At this point I have to add some remarks about testing: It is a
complete difference to shot a test image of "nothing" (cap on i.e.
simple dark image) or of a real lit object. In the second case
there forms an image which mostly "covers" the noise. If you take a
dark image only you simply will see the noise alone. Compare the
histograms of two of such images. They are very different, so a
comparison isn't truly fair that way.
Many images in astro photography are more similar to the dark image
testing. There will be only few spots in it to form an image:
stars. The huge rest is black (or in most cases unfortunately)
artificially brightened sky. Here you will see the noise rather
clearly cos you are digging in the lowest parts of your histogram.
This also is the cause, why my D 30 performs best when taking
images during twilight. As long as there is enough image signal
noise is secondary.
Another aspect: The tests were performed at ISO 100. This certainly
would not be the choise for the astrophotographer. He will look for
the highest sensitivity and use the fastest lens (I do with the
1,8/200 myself) to penetrate the skies as deep as possible. But as
we know noise really can get worse at higher ISO. How abot some
serious astro testing there with night sky images at ISO 800 and
1000?
I would like to get hands on a D 60 but won't have the chance so soon.
So could someone please do this for me and give us some more clues
about the real astro abilities of the D 60?

Rudolf, I started a more relevant thread here:
http://www.dpreview.com/forums/read.asp?forum=1019&message=2589322

Results so far seem to confirm that the D60 does indeed have significantly less noise than the D30. Please read through the whole of that thread. I think you'll find the discussion much more relevant.