Author | Thread |
|
10/19/2003 02:45:02 AM · #1 |
Do smaller apertures produce less or more grain? Does it matter what aperture I use?
Message edited by author 2003-10-19 02:45:24.
|
|
|
10/19/2003 03:36:40 AM · #2 |
Smaller apertures don't affect 'grain' or colour noise in digital cameras. The noise is related to ISO speed, the pixel pitch of the sensor (pixels/unit area), and the operating temperature of the camera. Noise can be more noticeable when using larger apertures, as it is easier to see noise in an evenly toned, blurred background than on something more in focus, which is usually more textured and/or busy. Smaller apertures will, however, accentuate dirt and dust in the system, as smaller apertures are forcing the light to take a more perpendicular path to the sensor, and hence dust particles and such will cast a sharper, more defined shadow on the sensor. This is more of an issue for those of us using interchangeable lenses where dust can make its way to the sensor quite easily. Dust on the front element won't cause anything more than a reduction in sharpness and contrast, even at the smallest apertures and under the dirtiest of conditions.
Yes, I should be out dancing with women...
Message edited by author 2003-10-19 03:49:18. |
|
|
10/19/2003 03:55:09 AM · #3 |
Thanks for the very interesting explanation James.
I was looking forward to find a good and precise answer like yours in this thread.
I knew about ISO speed only and I just suspected that aperture was not relevant in a direct way.
Do you know also how and why temperature affect noise on digital cameras? |
|
|
10/19/2003 08:44:33 AM · #4 |
Heat generates electrons in the pixels just like light does. More heat generates more pixels. CCD's used for scientific or astronomical use are usually cooled to well below freezing to reduce this so-called "dark noise". (But don't put your camera in the freezer! You'll risk condensation which would cause even more problems.)
Higher ISO speed is accomplished by amplifying the signal from the sensor. The noise is amplified as well, which is why noise is more apparent at higher ISO. |
|
|
10/19/2003 08:50:26 AM · #5 |
I believe that aperture can increase the noise level in a roundabout way... Smaller apertures will increase the exposure time. Some longer exposures will be more noisy by nature on digital cameras. It's generally not a problem at exposures under 1" in my own experience though.
|
|
|
10/19/2003 11:27:48 AM · #6 |
Wow. Thanks. Interesting stuff. Can't using the wrong ISO speed cause grain, too? Using a 100 when a 400 should be used? Or would I just get grain anyway from the 400 ISO? |
|
|
10/19/2003 11:31:56 AM · #7 |
the higher the ISO the more grain you will see
you should try to use a low ISO for nice crisp images - and only adjust it up if the light requirements - force you too.
or stick with a low ISO and use a tripod
for action shots an ISO of 400 would work a bit better than 100
anything over 400 and you will definately see grain.
soup |
|
|
10/19/2003 11:34:42 AM · #8 |
Originally posted by dsidwell: Wow. Thanks. Interesting stuff. Can't using the wrong ISO speed cause grain, too? Using a 100 when a 400 should be used? Or would I just get grain anyway from the 400 ISO? |
Using 100 when 400 is indicated should give you a much longer exposure (or underexposure), probably contributing to the heat-related noise.
I've noticed using exposure compensation produces grainier images -- perhaps from forcing longer exposures for a given ISO? |
|
|
10/19/2003 11:52:53 AM · #9 |
Originally posted by GeneralE:
Originally posted by dsidwell: Wow. Thanks. Interesting stuff. Can't using the wrong ISO speed cause grain, too? Using a 100 when a 400 should be used? Or would I just get grain anyway from the 400 ISO? |
Using 100 when 400 is indicated should give you a much longer exposure (or underexposure), probably contributing to the heat-related noise.
I've noticed using exposure compensation produces grainier images -- perhaps from forcing longer exposures for a given ISO? |
In my experience with both the10D and the Nikon 995, I always get less noise at lower ISO, even given the longer exposure necessary. I've experimented quite a bit with very long night exposures, and this has always been the case.
With regard to increased grain in shots using exposure compensation, if you compensate to the negative side (darker) then increase brightness in post-processing, you will have excessive noise in the darker areas of the image, as a result of "increasing the signal gain" through software. My personal rules for the smoothest possible images are:
1.) Use lowest ISO compatible with the required shutter speed
1a.)Use widest aperture compatible with desired DoF (more light, see above)
2.) Expose properly, don't try to bring up exposure excessively in software
3.) Under very difficult circumstances, e.g. night photography, take multiple exposures and "stack" in PhotoShop to reduce random noise (works wonders, but only for non-moving subjects...)
|
|
|
10/19/2003 11:57:46 AM · #10 |
Originally posted by kirbic: 3.) Under very difficult circumstances, e.g. night photography, take multiple exposures and "stack" in PhotoShop to reduce random noise (works wonders, but only for non-moving subjects...) |
thats a good idea
the ISO is the sensitivity to light - as that is increased you will always get more grain independant of the apeture and shutter.
other factors contribute to noise as well - but a lower ISO will result in less grain/noise.
soup |
|
|
10/19/2003 11:59:09 AM · #11 |
Those are all good hints. I especially want to try #3!
I mostly end up shooting with mostly automatic settings on my camera, but have been gradually been learning how to "trick" it into doing things for which there's no specific control available. |
|
|
10/19/2003 02:59:22 PM · #12 |
I've wondered this a few times.
The reciprocity of aperture, shutter speed and ISO mean that the exposure will be the same when you say double one and halve the other, but I am curious about how it impacts the noise.
So, going from ISO 100 at 1/250s to ISO 400 at 1/1000s for example (I think that's right, but regardless - higher ISO, shorter shutter speed) for the same exposure - do you get the same amount of noise ?
Don't know if the noise build up is linear - I suspect it isn't and as such, you might actually get _less_ grain at ISO 400 for the same overall amount of light - because the sensor is on for a shorter period of time, it doesn't warm up so much etc.
|
|
|
10/19/2003 03:21:25 PM · #13 |
Another consideration is heat noise from the battery, so the longer exposure required by the lower ISO may suffer from that, too. I recently bought the battery grip for the 10D...maybe moving the power source outside the main camera body might reduce noise somewhat. I wonder if anyone's done a comparison. |
|
|
10/19/2003 03:25:50 PM · #14 |
Originally posted by jimmythefish: Another consideration is heat noise from the battery, so the longer exposure required by the lower ISO may suffer from that, too. I recently bought the battery grip for the 10D...maybe moving the power source outside the main camera body might reduce noise somewhat. I wonder if anyone's done a comparison. |
Some samples from the D60, no battery grip. At ISO 100 the battery heat became noticeable once the exposure time got up to about 5 minutes - though that was after prolonged usage.
You can see the red blooming on the side.
I would have gone to a higher ISO but I wanted the longer exposures as a test for some star trail shots - seems they aren't possible with DSLRs yet...
//www.pbase.com/gordonmcgregor/360bridge
|
|
|
10/19/2003 05:24:05 PM · #15 |
with film anyhow - the only thing that would add noise would be an ISO change.
the exposure could be the same at a higher ISO - but noise would be higher.
i only have a Canon AL1 mainly a manual camera ( for film )- so battery is not an issue - in vermont - i think maybe in digital the battery heat will be dissipated quickly enough not to be a factor ;} - though that is new to me in the Dslr world.
soup
|
|
|
10/19/2003 08:25:12 PM · #16 |
With film, higher sensitivity comes from larger silver halide grains that collect light faster. The grains in faster film are large enough to be seen when the image is enlarged. Shutter speed has no effect.
Digital sensors are slightly sensitive to heat as well as light, so a longer exposure results in more noise. But the important thing is the signal to noise ratio. A exposure at a slower shutter speed has a bit more noise, but it also has a lot more light from the scene (unless it is very dark), so there is a lot more signal than noise. Using a fast shutter speed will result in less noise, but also less light; the ratio of signal to noise will be higher. A higher ISO will amplify the amount of light from the scene, putting it in the 0-255 range stored in the image. But a higher ISO will also amplify the noise, making it visible in the image.
Another type of noise is "fixed pattern noise", caused by variations between the pixels in the sensor. This isn't affected by temperature, but does increase with longer exposure times. It is most apparent in exposures above 1 second. |
|
|
10/19/2003 08:34:42 PM · #17 |
interesting, but still
the size of the hole has no effect on noise
soup |
|
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: 08/28/2025 07:48:43 PM EDT.