Wow, a long thread going on, well, it will take alot of my time to read all the posts but I wish I have time to read them all. anyways it will be a great place to learn.

Actually that looks like a fairly standard iso 6400 image I think. 6400 is really pushing any camera even the D700. The only current Nikon camera that can handle it is the D3S. Have you tried Noise Ninja software? Its really good within reason it ca help remove a lot of noise.

I'd like to see them. Do you have a link to a gallery?

jonnyapple said:

Long exposure noise reduction is really just a name for dark current subtraction.

Your noisy images are normal and nothing to worry about, just be thankfull that your exposures are only 30 seconds.

There are two noise issues that need to be looked into.
When you do astronomical photography, your exposure times are up in the high hundreds os seconds, 300, 600, 900. This introduces huge numbers of noisy pixels (thats why most astronomers use canons, but thats another issue).

Astronomers will take dark frames (cover the lens cap) and image for exactly the same length of time. Normally 10-15 darks are taken on average. This image records all the following noise data.
1) Hot pixels.
2) Amp glow (heating effects due to camera electronics. Long exposures in summer are noiser then the same length of time in winter!!!)
3) Readout noise.

we then take bias frames. These are images of the readout noise which appear in EVERY photo. These are the fastest shutter speed your camera does (with lens cap on) again between 10 - 15 are normally taken.

Then you have to think about flat frames (dust off images)

So your 30 second image contains the following data
1) Light image (what you can see)
2) Hot pixel noise
3) readout noise.

The noise reduction in the nikons only address the hot pixel noise not the readout noise. Plus the fact their raw images are not true raw!!!

You can use software to combine your images to remove the noise and end up with a true noise free image.

The reason for multiple dark & bias frames is to remove the random noise from high energy particles striking the sensor.

Mac.

If i get time, i'll post examples

Whoa, I wouldn't consider myself an expert, Pierre. I'll take the compliment, though. Thanks.

Long exposure noise reduction is really just a name for dark current subtraction.
http://en.wikipedia.org/wiki/Dark_current_(physics)
http://en.wikipedia.org/wiki/Dark_frame_subtraction

JorPet is right about why the NR works. It will always be the same pixels.

If you think about it, if the same pixels did not get hot during the second "dark" image that the camera uses to correct the original, then it simply wouldn't work. It would make sense that given pixels will always do the same thing under the same condition which allows long-exposure noise reduction to work.

Thanks Jonny and Niko, it is nice to get expert opinion, I surely feel better to know that.

That is normal. One solution is to use lightroom. It's demosaic routine is smart enough to remove them automatically if you shoot raw. I think long exposure NR will take care of it, too, as was mentioned above.

Here is one, it was taken with 30" at ISO 6400

Lots of hot pixels, I took many shots and they show the same hot pixels, at the same location and with the same color. Many are at the top left.

There are more in the Photo-A-Day.

I agree with NikoDoby on this. How about posting a pic Pierre then we can tell for sure. I often shoot on a tripod at L 0.3 to iso 200 at 30 seconds or bulb and usually get what I want out a scene. Thats strange how you have problems with iso 1000. Noise is natural of course with higher iso ranges, but within reason on the D700 I think. Please post a pic for us!

How else is he going to really be satisfied if he doesn't check it agains a separate copy?

There would be a point in returning the camera only if you get dead pixels (dark or bright pixels) at any iso. If not then do what NikoDoby already suggested and turn on the long exposure NR.
I have it on all the time as it will save you a lot of trouble in PP later and I believe that the long exposure NR (not the ordinary in camera noise reduction that some cameras have in addition) in camera will get you a better result than you can achieve in PP as the camera knows exactly what pixels get hot and will remove those values from the final image while NR software will not have that information. Correct me, if I'm wrong.

What?!? That makes no sense Photofan. Noise at long exposures is perfectly normal. How is returning the camera for one that will do the exact same thing going to help? Oh yeah I'll be able to buy an open box D700 for the price of a D300S and NOT have to pay $7000 to get that rebate :^)

Seriously though Pierre just post up a photo of the "problem" and let us take a look at it. If we see something out of the ordinary like Bigfoot we'll let you know it's cause for concern.

Here's an explaination of what a stuck pixel is. (scroll to the bottom to see the long exposure sample)
http://www.dpreview.com/learn/?/key=noise+reduction

Pierre said:
a brand new $2500 camera that has the reputation of seeing in the dark.

No that reputation would belong to the $5000 dollar D3S :^)
The D700 is good for low light but it can't see in the dark like the D3S can.

If this is a brand new camera and the issue really bothers you, I'd return the camera for another one. There's no other way to convince yourself whether you got a bad copy or not.

kanuck said:
Did you try the advice on here and turn on these options?

The NR was off and I surely could turn it on, map the bad pixels or just Photoshop them out and surely, I will not make a habit of shooting with the cap on or in total darkness.

I am still wondering if it is really normal to get so many hot pixels on a brand new $2500 camera that has the reputation of seeing in the dark. These overacting pixels are not random neither; it is always the same ones that come up and always in the same color, either red, blue or white.

Surely, I do not expect the camera to catch on fire or explode but I still would like to know the effect of long exposures on the lifespan of a pixel, so that I could avoid behavior that would cause damages.

If on the other hand, I got a bad sensor copy or I am just being picky, I need to know now so I can get it fixed, replaced or simply avoid making a foul of myself.

NSXType-R said:
Can you set your camera for longer than 30 seconds other than bulb?

I am still new to the D700 but I think the answer is no. You would need to time yourself and use bulb for longer exposures.

I am not sure neither what x250 does below 30" nor what speed it is actually shooting.

Interesting post. I have never experienced this problem before even though I shoot a lot of night exposures especially long after the sun goes down. I never use long exposure or noise reduction either. I rarely go up as high as 6400 unless its an emergency although you mentioned that even at iso 1000 you got those cooked pixels? Did you try the advice on here and turn on these options?

NR and Long exposure option turned on should fix this but, if you don't need them on don't forget to turn them off because I have found they give my images more noise similar to that found in my brother's canon 5d Mark II but not as bad.

Can you set your camera for longer than 30 seconds other than bulb?

Almost all cameras have this. Just turn on long exposure noise reduction and your all set.

The sensor naturally heats up during long exposures. No big deal. It's not going to blow up in your face or anything like that if you don't stop. Just turn on in camera noise reduction NR. But for for a 30 second exposure the camera will "lock-up" for about 30 more seconds while the image is processed for noise. If you don't want to wait an entire minute for one photo to show up on your LCD screen then you can also map out the noise later with software. But it's easiest to just let the camera do it for you.

PB PM said:
pixels get hot after long expsoures

Do you mean that I should had given time for the sensor to cool after a 30 seconds exposure?
The Camera is 3 weeks old, and about 1500 activations. I would not have expected so many hot pixels in a new camera.
Do you mean that these hot pixels will return to normal after some cooling time? Or they are plain dead?

That is normal, pixels get hot after long expsoures. That is why Nikon includes Long Exposure Noise Reduction in camera. After the photo is taken it will take another exposure of the same length, but with the shutter closed so it can eliminate hot pixels from the image.

Last weekend for the fun of it, I was shooting in the dark, trying to capture a feeble moon reflection on a nearby stream.

I cranked the ISO to 6400, mount my new D700 and 24-70 on the tripod and set the speed to 10 seconds, then 20, then 30. Back home I looked at the result and found the pictures peppered with red and blue dots, all at the same locations and same color on each pictures and I stopped counting after 30 of them.

I then tried to shoot 10 seconds with ISO 1000 and the lens cap on, same overacting pixels but this time a bit fainter than at ISO 6400.

Is there something wrong with my sensor or is this normal?