Nikon D200: Long Exposure noise

[mikelong]

I'm seeing considerable noise from my D200 from exposures lasting more than two seconds - exposures running six minutes are basically unusable. I'm using the long exposure noise reduction setting but the results are unsatisfactory. Is this normal? Is there another digital camera available that provides better results? I'd like to shoot exposures up to one hour or more (battery power loss is another issue, for the time being I carry four Lithium-Ion batteries to deal with it).

Here's an example from a recent trip to southern Utah, click on the image to see the noise. I've cleaned up the image considerably, which can be seen here: http://www.pbase.com/mikelong/utah



Thanks
Mike

[dang]

Refer to this thread concerning dead pixels: http://forum.pbase.com/viewtopic.php?f=13&t=37534
The only thing I can think of is the purple blob is lens flair, similar to when shooting IR, but from where I haven't a clue.
Hope you sort it out.

[prinothcat]

Refer to this thread concerning dead pixels: http://forum.pbase.com/viewtopic.php?f=13&t=37534
The only thing I can think of is the purple blob is lens flair, similar to when shooting IR, but from where I haven't a clue.
Hope you sort it out.

Looking at the light and shadow I bet he moon is in that area....
As for the noise issue, I'd have to look but I don't have inordinately bad noise on long exposures with a D200.

and

are both long exposures. I wonder if the large range of contrast in the OP 's image has something to do with it.

[mikelong]

Looking at the light and shadow I bet he moon is in that area....
As for the noise issue, I'd have to look but I don't have inordinately bad noise on long exposures with a D200.
I wonder if the large range of contrast in the OP 's image has something to do with it.

The moon is completely behind me in this shot, well on it's way to setting in the west. And why would it be magenta? This can't be normal...

[dang]

The moon is completely behind me in this shot, well on it's way to setting in the west. And why would it be magenta?

It could easily be that since there is such a high contrast difference, the light in the lower half is bouncing upward producing the spot. As to why the coloring, probably due to the lens coatings. If you were using a filter, it could also play part. For
night scenes, it's usually best to remove them.

[mikelong]

It could easily be that since there is such a high contrast difference, the light in the lower half is bouncing upward producing the spot. As to why the coloring, probably due to the lens coatings. If you were using a filter, it could also play part. For night scenes, it's usually best to remove them.

Thanks Dang. No filter was used for the five minute exposure. I just ran a 4-minute test similar to that proposed in the thread, no filter was used. Click on the image to see the lousy result.
Mike

[dang]

Hi Mike,
Think I'd contact Nikon... Makes me wonder if they had a bad batch of sensors,
or some other problem. Let us know what the outcome is, if you don't mind.
Thanks,
tom

[mikelong]

Hi Mike,
Think I'd contact Nikon... Makes me wonder if they had a bad batch of sensors,
or some other problem. Let us know what the outcome is, if you don't mind.
Thanks,
tom

Hi Tom, here's the response I just received from Nikon, which is totally "out there". I believe he actually thinks my test shot with the lens cap on is of the night sky, and he definitely didn't look at the EXIF data:

"This resembles a light leak, but that's not what causes this. This looks like a CCD that is generating a lot of heat from prolonged charging. You may notice that CCD cameras designed for telescopes have heat syncs on them, and some even have a cooling fan or a 'coolant' system, and this is the reason. Some telescope cameras use what's known as a Peltier coolant device to carry heat away from the sensor.

In your case, the amplifier in the sensor is cranked all the way up to an ISO of 1600, combined with an 8+ minute exposure and a very dark subject, and this is the result. You may want to lower the ISO way down to 400, and wait for the ambient temperature outside to drop more to around 40ºF (no lower than 35ºF) for space shots. Turn off the VIVID mode, and leave off the Noise Reduction (it erases stars).

Use the fastest (largest aperture) lens you have to limit the time needed to collect light. I notice from the shooting data, you have the Aperture set to f/22 for the shot with the most purple. Open that lens all the way up, you can do this in Manual mode.
Allowing the camera to cool down in between shots may help too. This is the best advice I give you for the best chance with this camera and technology, but keep in mind, it is not designed for that function.
There are many differences between CMOS and CCD sensors. CCDs are more costly to produce, but generally produce a much better image. CCDs use more power than CMOS sensors and as a result, will generate less heat. This is why you may not have this problem at this scale with a CMOS camera.

Thanks for using Nikon products!"

Me again...these shots were both shot wide open for between 250 and 360 seconds at ISO 400...!
Mike

[dang]

Hi Mike,
I know it's not funny, but you can't help but chuckle... This really reminds me of Canon's responses!
I think I'd resend it, and include the exif in the wording to avoid his obvious confusion. While he makes some valid points,
they're in left field in your situation. Maybe try doing a series of shots starting a 15 seconds, and double them each time.
At least you'd get an idea how slow you can expose. If it's starting at 15, or 30 seconds... they'll be hard to explain away
by a heated sensor. Even a minute, or two shouldn't show so many dead pixels. Or at least my Canon bodies haven't so far.

Wonder how many others are getting this with the D 200/300's? Might be worth searching for...

[prinothcat]

If someone will post a link to the test, I'll run it on my 1 year old D200. Other input, I've seen the purple fog before Ken Rockwell mentions it here bit not in reference to the D200, and google tosses up this page of results when one searches CMOS charge leakage.

[dougj]

Here is a mention of the purple haze with long exposures an a D200.
http://photo.net/bboard/q-and-a-fetch-msg?msg_id=00LsBA

Further on in the thread a user mentions a program called Startrails, with a link, written by an astrophotographer. This sounds interesting and might be worth a look.

[prinothcat]

Further on in the thread a user mentions a program called Startrails, with a link, written by an astrophotographer. This sounds interesting and might be worth a look.

Too bad it was only written for winblows.

[dougj]

Too bad it was only written for winblows.

It might be useful to the OP, I believe he is using PS CS for Windows.

[mikelong]

Thanks for the links. I sent Nikon the real EXIF data since the "tech" seems to believe I was shooting at f/22, IS0 1600 for 8+ minutes, and that this would explain the problem. Where they are getting their data I have no idea since both images were shot wide open for three to four minutes at ISO 400. Either way, I'm occasionally getting digital noise or "hot" pixels on exposures as short as three-quarters of a second.

[mikelong]

Hi folks, here's the latest from Nikon regarding this issue. They requested an NEF file so I shot another test image (ISO 200, f3.5, 310 seconds). There appears to be less noise because the image is almost completely black, but after playing with levels you can see the noise is still there (not to mention plenty of magenta discoloration) if you click on the image below.
Mike

Hi Mike
If you look at the last image you sent the noise is much less apparent than previous images.
Nikon cameras are not made specifically for Astrophotography.
My suggestion would be to use as low an ISO as possible.
In tests with our D200 I did 5 minute exposures at 100, 200, 400, 800 ISO.
I only started to see noise at ISO 400.
You can always send your camera in to be evaluated but it is performing as expected.

Jeff

Thanks for using Nikon products!