| Author | Thread |
|
|
08/01/2007 02:39:27 PM · #1 |
| I have recently started shooting night shots using long exposures as of right now I have no remote so I am stuck to the Max 30 sec exposure. I has taken some nice pictures (inRaw) but the noise it almost over wellming. I know I can take some boise out in the RAW conversion but what else will giv me the super clean night shots I see all over this site. Thanks for the help. |
|
|
|
08/01/2007 02:41:44 PM · #2 |
| Do you have noise reduction on your camera I have it on the d50 and it seems to help but takes forever. |
|
|
|
08/01/2007 02:43:48 PM · #3 |
neat image or noise ninja seem to be the go to programs for this site. I personally use neat image, they have a free download, check it out it makes a world of difference.
|
|
|
|
08/01/2007 04:39:57 PM · #4 |
A common method is dark frame subtraction. This involves taking another shot of the same length as the original but with the lens cap on. So, you end up with a frame of just the hot pixels and noise.
Then go into PS and place the dark frame as a layer over the original shot, use 'difference' (I think) as the blend type. The object is to cancel out the noise and hot pixels from the original layer.
Of course, in-camera noise reduction does almost exactly the same thing. |
|
|
|
08/01/2007 04:46:52 PM · #5 |
There's a technique involving multiple exposures. For non-moving subjects, take three or more shots that are exactly the same. Use the first exposure as background, the 2nd one as a layer on top with 100/2=50% opacity, the 3rd one with 100/3=33% opacity. The more shots you take, the less noise you end up with. Doesn't fix hot pixels though.
|
|
|
|
08/01/2007 04:59:44 PM · #6 |
| low ISO rating would be better for night photography then high ISO. Higher the value would result in high noise. |
|
|
|
08/01/2007 05:05:02 PM · #7 |
I would agree with the lower ISO but after seeing some shots taken by  larus i have to wonder. He uses ISO 1600 on a lot of his shots and they all come out awesome. Not sure how he does it.... |
|
|
|
08/01/2007 05:07:17 PM · #8 |
| Incidentally, if your shutter speed is slower than about 1 minute, then you're going to start seeing star trails. Which means you're going to have to get a tracking mount. Which is a whole different story. |
|
|
|
08/01/2007 05:08:45 PM · #9 |
But if noise is random, how does dark fram subtraction work, if the actual noise is in a different place from one shot to the next.
I can see it working for hot pixels, those are usualyl stationary. |
|
|
|
08/01/2007 05:12:41 PM · #10 |
Originally posted by LanndonKane:
But if noise is random, how does dark fram subtraction work, if the actual noise is in a different place from one shot to the next. |
Dark frame subtraction is for dealing with hot pixels. |
|
|
|
08/01/2007 05:13:12 PM · #11 |
Originally posted by LanndonKane:
But if noise is random, how does dark fram subtraction work, if the actual noise is in a different place from one shot to the next.
I can see it working for hot pixels, those are usualyl stationary. |
The cause of the noise, which occurs on the quantum level (excited atoms firing out photons) has an element of randomness to it, but the effect at the level of the sensor is pretty predictable. As long as conditions don't change radically (length of exposure, ambient temperature, etc.), the noise will look similar enough that a simple subtraction will show a good result. Many astronomers will shoot multiple dark frames at the same exposure length and "average" them in a processing program, and then use that composite as their dark frame.
I have shot night shots with and without noise reduction turned on in my 30D and the outcome is not all that different. I don't think the incamera noise reduction is all that effective. I'll post a few examples if I can scrounge some up. |
|
|
|
08/01/2007 05:44:43 PM · #12 |
| D'oh, my bad, I should have realized that quantam mechanics came into play :P |
|
|
|
08/01/2007 05:58:39 PM · #13 |
| What about using bulb to get a longer exposure, hence being able to shoot at lower ISO? (haven't tried it yet, being a brand-new dslr owner, so might not know what I'm talking about :) ) |
|
|
|
08/01/2007 06:22:15 PM · #14 |
He can, I dont think he wants too without a cable release for some reason. Maybe he's worried about camera shake.
If this is the reason, you can avoid camera shake or mirror slap by holding a hat or piece of dark cardboard in fron the lense while you open and close the shutter, and wait for the camera to settle before removing the obstruction. |
|
|
|
08/01/2007 06:59:10 PM · #15 |
Originally posted by LanndonKane:
He can, I dont think he wants too without a cable release for some reason. Maybe he's worried about camera shake.
If this is the reason, you can avoid camera shake or mirror slap by holding a hat or piece of dark cardboard in fron the lense while you open and close the shutter, and wait for the camera to settle before removing the obstruction. |
Yup. This is what I'd do. Use this trick, lower ISO, and longer exposure. Using the camera's long-exposure noise reduction will automatically do the "dark frame subtraction" though the exposure will take twice as long because the camera needs to take the dark frame with the same parameters as the image.
There are two types of noise, random and fixed-pattern. The random noise, as strangeghost pointed out, is due to quantum effects including thermal noise and photon arrival statistics. This noise varies its pattern with each exposure and can be reduced by using multiple overlaid shots, as suggested by gloda. Fixed pattern noise is a function of the sensor and readout electronics, and is effectively removed by dark frame subtraction, since it has the same pattern in each shot. |
|
|
|
08/01/2007 07:01:06 PM · #16 |
Originally posted by wizardry:
I would agree with the lower ISO but after seeing some shots taken by larus i have to wonder. He uses ISO 1600 on a lot of his shots and they all come out awesome. Not sure how he does it.... |
larus shoots with a Canon 5D. There is arguably no better camera for this purpose (low-light shots at high ISO), period... though the Canon 1-series cameras are on roughly equal footing. |
|
|
|
08/01/2007 07:08:11 PM · #17 |
Originally posted by kirbic:
Using the camera's long-exposure noise reduction will automatically do the "dark frame subtraction" though the exposure will take twice as long because the camera needs to take the dark frame with the same parameters as the image. |
In bulb mode, does the camera know how long to shoot the dark frame for? Or does it just default to 30 seconds? |
|
|
|
08/01/2007 07:13:13 PM · #18 |
Originally posted by jhonan:
In bulb mode, does the camera know how long to shoot the dark frame for? Or does it just default to 30 seconds? |
Regardless of the mode, the dark frame is always exactly the same length as the exposure. IOW, if in bulb you shoot a 4.5 minute exposure, the cam will then shoot a 4.5 minute dark frame.
Message edited by author 2007-08-01 19:13:38. |
|
|
|
08/01/2007 08:33:38 PM · #19 |
Originally posted by wizardry:
I would agree with the lower ISO but after seeing some shots taken by larus i have to wonder. He uses ISO 1600 on a lot of his shots and they all come out awesome. Not sure how he does it.... |
a) you are seeing a 640 pixel version. Lots can be hidden at that size.
b) Larus shoots with f/1.4 and f/1.2 lenses (still L-glass) which decreases the time needed for the exposure. 30 seconds for him may be 2 minutes for someone else.
|
|
|
|
08/02/2007 10:13:16 AM · #20 |
| I have a lot of testing to do thanks for all the good info... |
|
|
|
08/02/2007 11:36:30 AM · #21 |
| Are you bumping up the exposure in RAW because the 30 second exposure isn't quite enough? If that's the case, that will REALLY increase the noise in a long exposure. I would definitely look into a remote release. I don't know about Nikon goods, but I don't think a basic remote would cost too much. |
|
|
|
08/02/2007 11:41:42 AM · #22 |
Originally posted by Telehubbie:
I don't know about Nikon goods, but I don't think a basic remote would cost too much. |
About US$20 |
|
|
|
08/02/2007 12:14:50 PM · #23 |
Here are two shots from an astro shoot last month. I know that astronomy shots are not exactly a real-world test of a camera's performance, but where else do you shoot 4.5 minute exposures?
 
Noise reduction Off....... Noise reduction On
3:28 am .......................... 3:10 am
Both Shots: July 15, Canon 30D, Tokina 12-24mm at 12mm f/4.5, ISO 800, 270 seconds, center crops at 100% (so you're looking at actual pixels). The NR-Off shot was taken about 18 minutes after the NR-on shot, and the sky was beginning to lighten with very early dawn glow (we were at 46 deg N lat.). Other than cropping, images are completely unprocessed, saved at JPG 92%.
For my money, the benefit gained by using in-camera noise reduction is not that significant - for astrophoto purposes. Much more beneficial is shooting multiple exposures of the same composition and stacking them in photoshop.
Message edited by author 2007-08-02 12:18:54. |
|
|
|
08/02/2007 07:54:58 PM · #24 |
| no I am not bumping up the exposure in RAW conversion usually I have enought light for a 30 sec to work... |
|
|
|
08/02/2007 08:53:19 PM · #25 |
Not only that, but the temperature of the sensor for the dark frame must be close to the image. In other words, if you are outside taking a long astrophoto shot, don't go indoors and take the dark frame. Warmer temperatures generate more noise. Likewise, longer exposures generate more heat, which generates more noise.
Originally posted by kirbic:
Yup. This is what I'd do. Use this trick, lower ISO, and longer exposure. Using the camera's long-exposure noise reduction will automatically do the "dark frame subtraction" though the exposure will take twice as long because the camera needs to take the dark frame with the same parameters as the image.
|
|
|
Home -
Challenges -
Community -
League -
Photos -
Cameras -
Lenses -
Learn -
Help -
Terms of Use -
Privacy -
Top ^
DPChallenge, and website content and design, Copyright © 2001-2025 Challenging Technologies, LLC.
All digital photo copyrights belong to the photographers and may not be used without permission.
Current Server Time: 08/04/2025 09:21:52 AM EDT.
