I was under the impression that my first photo example exhibits amp noise AND amp glow while the second photo only exhibits amp glow?
Image stacking and NR can help with amp noise but not amp glow?
Can a very long exposure damage a camera sensor
(43 posts) (10 voices)-
Posted 3 years ago #
-
(I've never heard the term "amp glow" used for what you're seeing in the corners there before, but I'll run with it.)
If ambient temps don't change, dark frame removal will cancel out 100% of "amp glow". It's a systematic error caused by charge building on the sensor independent of photons. 100% reproduceable, and therefore 100% cancelable.*
The noise pattern I see in the first photo but not the second photo looks characteristic of high-ISO noise, which is random noise amplified at higher ISOs. Random noise such as this can only be modeled and removed by its characteristics, not by negation. I think this random pattern of single pixels is what you're calling "amp noise" - which is not the term I'm familiar with for this type - this is random noise amplified, but not a result of the amps themselves. This is seen at high ISOs because in low-exposure situations (or areas) the Signal to Noise ratio is low, so the amps are amplifying a statistically significant amount of noise, but (as I said) that's not the amp's fault - they didn't create it - which is why it's not commonly called amp noise.
What you are calling "amp glow" is, I believe, more commonly called "amp noise" as it IS noise (thermal noise) caused by circuity adjacent to the sensor heating the sensor. This is an exacerbation of the latent dark current. All noise, when it comes down to brass tacks, is thermal noise - run your sensor at 0K and you won't see any of this. ;)
Thom Hogan covers this decently in all his Nikon guides. (Is he as verboten as KR?) Though he counts hot pixels as random noise wherein I (and many others) classify them as a systematic noise source as they are predictable (if you know the conditions (which is why I mentioned needing roughly consistent ambient temps for the two shots (photo and dark frame))), which is why Nikon's long-exposure noise reduction can be so successful against them.
*"100%" in this context meaning 100% /perceptually/ removed - as variations below the noise floor should be expected.
EDIT: But, yea, in direct response to your earlier question, post-camera software would be nearly inable to deal with what you're calling "amp glow" as it (the post-camera software) has no idea what is intentional and what isn't. In-camera, though, is a whole other story as, all-things-being-equal, the noise is perfectly reproduceable.
EDIT2: Doing a 5, 10, or even 30 minute exposure you're likely to get excellent results from dark-frame canceling. I brought this up waaaay up there because I'm not so sure after 4 hours one can expect as good of results - outside the artic I'm not sure there are many environments where the ambient temp is /that/ consistent over the needed 8 hours. (not to mention the fact you might not reach thermal equilibrium inside the camera case until quite a while into the first shot, while the second shot will be completely at that temp.)
EDIT3: Edit2 brought up another good point - when doing long exposures with in-camera long-exposure-noise-reduction it is very important that you allow your camera to acclimate to ambient temps before taking the shot. A camera still warm (or cool) inside after spending time in your climate-controlled house/car while the shot is being taken will have a different internal temp when the second (noise reduction) frame is being measured and will throw off the results.
Posted 3 years ago # -
So camera noise reduction can not reduce "amp glow"? Give me some in camera settings to try if you say it can. I usually see "amp glow" at around 5 plus minute exposure even with NR on.
Posted 3 years ago # -
"So camera noise reduction can not reduce "amp glow"?"
No - I'm saying amp glow is one of the two noises it CAN remove almost entirely.
Which camera, which noise reduction? You should have two options: (at least both my D700 and D200 do (shooting menu))
Long Exp. NR
(and)
High ISO NRLong Exp. NR does a dark frame removal (inversion addition, whatever), and should kill "amp glow" dead.
High ISO NR tries to kill those random pixels (not hot or stuck ones, but the low-level "grain" like noise) but at the cost of detail. This only affects JPG / TIFF IIRC. I leave it off anyhoo, and use neatimage or noise ninja if it is something I really want to attack. (EDIT2: They have the advantage of letting you do a context-sensitive attack on the random noise, which tends to do a more aesthetically pleasing job)
EDIT:
IIRC my D50 had the same two options, but it had an out-of-spec hot component inside and could actually remove _too much_ on Long Exposure NR and cause black spotting.Posted 3 years ago # -
Wow you guys are giving me a headache with all this talk. Less talk more action. Niko you have a few cameras just take one and let it expose something for a few hours and see what happens
I still think I am going to try the film way, do you think I would damage my old FG camera if it exposed for 8 hrs at app minus 10 to 15 C area. I just need to find a spot that have great view without the chance of cars and other lights damaging the picture. Maybe there would be a problem with frost on the lens (filter) Or I will have to wait to next spring
Pete
Posted 3 years ago # -
D70, D50, D80, D200, and on one rare occasion a D300 at a very long exposure. I've had "amp glow" even with NR on as I've already stated.
So on your D200 you completely eliminate "amp glow" by just using Long Exp. NR?Posted 3 years ago # -
Actually I prefer using my pinhole camera for long exposures Pete :^)
Posted 3 years ago # -
yes, completely remove it.
How long of demo shots do you want?
(I believe this will show up in the EXIF tags, no?)
Posted 3 years ago # -
I have a Nikon pinhole lenscap, could you expose stars with a pinhole lens and if yes which film speed should I use
Sorry for being stupid on this, but we are here to help right, with a bit of humor mixed in for good measurePete
Posted 3 years ago # -
Very slight amp noise seen @ 13 mins no noise reduction.
will do a longer NR shot next. Will upload tomorrow, my FTP client is on a thumb-drive in the truck and it's raining HARD. ;)
Posted 3 years ago # -
Came out well, 825 second exposure shows amp noise on edges and corners, 1095 second exposure with NR on shows none. Will upload to my site tomorrow.
Posted 3 years ago # -
With the D200?
Posted 3 years ago # -
yes
Posted 3 years ago # -
Long timed exposures should not damage a sensor. However, pointing the camera at the sun or other really strong light source could harm the sensor. If you are using the live view mode and you point it at a very strong light source there is a risk of doing extensive damage to the censor and other electrical parts in the camera. On a side note, the AF- assist illuminator self-timer lamp/ red eye- reduction lamp (same thing) is prone to overheating especially when shooting on Continious mode. All that happens is it will shut done when it overheats. It functions again once it cools down. Keep shooting those long timed exposures! Magical :)D
Posted 3 years ago # -
mmm k.
I don't see the amp noise on my PC on the 18 min shot. Only on the camera screen.
I'm about to leave the hotel and drive for a while - I'll see if I have enough battery power to do a much longer test while I drive. (My AC adapter is @ home :( ) If not it will have to wait until I get to the next hotel and charge up.Posted 3 years ago # -
Long exposure noise reduction off.
Long exposure noise reduction on.Quite black and white improvement, check the metadata for exact setup. The one with NR on is actually a longer exposure - and should have experienced more amp noise.
http://www.flickr.com/photos/43765368@N07/
Figured - what the hell - why do a looooong exposure when I can just crank up the ISO and do a long exposure to demonstrate amp noise and how easy it is to fix.
Posted 3 years ago # -
Your "long exposure noise reduction on" photo doesn't show up soap?
Posted 3 years ago # -
yes I'm kidding with you
Posted 3 years ago #
Reply
You must log in to post.
