2007-10-24 10:15:58
Are digital camera exposures affected by reciprocity failure the way that film exposures are?
Thanks for any advice.
Thanks for any advice.
#1
Are digital camera exposures affected by reciprocity failure the way that film exposures are?
Are digital camera exposures affected by reciprocity failure the way that film exposures are?
Thanks for any advice.
Luis Ortega wrote:
Are digital camera exposures affected by reciprocity failure the way that film exposures are?
Thanks for any advice.
No its not like that its called noise.
To rectify this you should use something like Noise Ninja (Google it up). Demo version should show you how the problem is resolved.
Rob wrote:
Luis Ortega wrote:
Are digital camera exposures affected by reciprocity failure the way that film exposures are?
Thanks for any advice.
No its not like that its called noise.
To rectify this you should use something like Noise Ninja (Google it up). Demo version should show you how the problem is resolved.
Reciprocity failure in film is not related to noise in any way, other than that both are most evident at low light levels. They involve different physical processes.
What you are looking at is called NOISE.
On Thu, 25 Oct 2007 03:02:11 +1000, Rob
wrote:
What you are looking at is called NOISE.
He didn't ask about noise, or say he was "seeing" anything. Read the question again:
"Are digital camera exposures affected by reciprocity failure the way that film exposures are?"
The answer is "no", noise doesn't come into it.
Are digital camera exposures affected by reciprocity failure the way that film exposures are?
Thanks for any advice.
Are digital camera exposures affected by reciprocity failure the way that film exposures are?
Thanks for any advice.
Thanks everyone.
I think that I understand the difference between noise and reciprocity failure.
If you get digital noise when photographing in low light, will increasing exposure time reduce the noise or just add to it?
I am thinking primarily about long exposure night photography and doing things like traffic light trails in cityscapes or doing light painting or light tracing techniques.
Thanks everyone.
I think that I understand the difference between noise and reciprocity failure.
If you get digital noise when photographing in low light, will increasing exposure time reduce the noise or just add to it?
I am thinking primarily about long exposure night photography and doing things like traffic light trails in cityscapes or doing light painting or light tracing techniques.
"Luis Ortega" wrote in message
Are digital camera exposures affected by reciprocity failure the way that film exposures are?
Thanks for any advice.
Luis Ortega wrote:that
Are digital camera exposures affected by reciprocity failure the way
film exposures are?
Thanks for any advice.
No its not like that its called noise.
To rectify this you should use something like Noise Ninja (Google it up). Demo version should show you how the problem is resolved.
On Thu, 25 Oct 2007 03:02:11 +1000, Rob
wrote:
What you are looking at is called NOISE.
He didn't ask about noise, or say he was "seeing" anything. Read the question again:
"Are digital camera exposures affected by reciprocity failure the way that film exposures are?"
The answer is "no", noise doesn't come into it.
--
John Bean
"Luis Ortega" wrote:
Thanks everyone.
I think that I understand the difference between noise and reciprocity failure.
If you get digital noise when photographing in low light, will increasing exposure time reduce the noise or just add to it?
I am thinking primarily about long exposure night photography and doing things like traffic light trails in cityscapes or doing light painting or light tracing techniques.
If you have to deal with reciprocity failure, the effect is that when you stop the lense down one fstop, instead
of increasing the exposure time by 2x, it will need to
be more than 2x.
That doesn't happen with digital. But with digital when a longer the exposure time is needed because there is so little light, the more variation there is in sensing of
that light, and that is noise. Interestingly, other
kinds of noise might actually have a reduced effect due
to there being enough time that a average is smoother or more consistent.
That means your results will depend on just how you deal with low light photography. For example, several
distinct exposures can be added, or they can be
averaged, or just one long exposure can be made. The
effects are different. It depends on whether you want
to see a persistent low level light, or an intermittent
higher level light..."Luis Ortega" wrote in message
Are digital camera exposures affected by reciprocity failure the way that film exposures are?
Thanks for any advice.
Marvin wrote:I'm not the only one who read your response differently than you seem to have meant.
Rob wrote:
Luis Ortega wrote:
Are digital camera exposures affected by reciprocity failure the way that film exposures are?
Thanks for any advice.
No its not like that its called noise.
To rectify this you should use something like Noise Ninja (Google it up). Demo version should show you how the problem is resolved.
Reciprocity failure in film is not related to noise in any way, other than that both are most evident at low light levels. They involve different physical processes.
Do you think I should have spelt it out better?
I said NO.
What you are looking at is called NOISE.
Of course you are looking at different physical properties.
"John Bean" wrote in message
On Thu, 25 Oct 2007 03:02:11 +1000, Rob
wrote:
What you are looking at is called NOISE.He didn't ask about noise, or say he was "seeing" anything. Read the question again:
"Are digital camera exposures affected by reciprocity failure the way that film exposures are?"
The answer is "no", noise doesn't come into it.
You are correct. The answer is "no", since reciprocity is a chemical behavior of the film's photosensitive crystal , i.e., losing sensitivity as a function of the time of exposure. There is "no parallel" in the digital paradigm! Noise has nothing to do with it.
Rob wrote:
Marvin wrote:I'm not the only one who read your response differently than you seem to have meant.
Rob wrote:
Luis Ortega wrote:
Are digital camera exposures affected by reciprocity failure the way that film exposures are?
Thanks for any advice.
No its not like that its called noise.
To rectify this you should use something like Noise Ninja (Google it up). Demo version should show you how the problem is resolved.
Reciprocity failure in film is not related to noise in any way, other than that both are most evident at low light levels. They involve different physical processes.
Do you think I should have spelt it out better?
I said NO.
What you are looking at is called NOISE.
Of course you are looking at different physical properties.
I think that I understand the difference between noise and reciprocity failure.
If you get digital noise when photographing in low light, will increasing exposure time reduce the noise or just add to it?
While some punctuation and proper spelling may have helped,
"No. It's not like that. It's called noise."
the meaning was quite clear, if read without an agenda!
Johan W. Elzenga johan<<at>>johanfoto.nl Editor / Photographer http://www.johanfoto.com
In article <tJRTi.346$>,
"Luis Ortega" wrote:
I think that I understand the difference between noise and reciprocity failure.
If you get digital noise when photographing in low light, will increasing exposure time reduce the noise or just add to it?
Longer exposure times will add to the noise; you'll see increased effects from thermal noise and from differences in sensitivity within the individual sensing elements of the CCD.
--
Photography, kink, polyamory, shareware, and more: all at http://www.xeromag.com/franklin.html
I am trying to understand this concept.
Going away from Reciprocity and turning this to noise control in multiple exposures. Or if you feel I should I could start a new thread about this??
So if longer exposure times will add to the noise, then multiple exposure should also add to it due to the build up of detail? Even before I start to experiment with multiple exposures I would like to be able to understand what I'm supposed to be facing here. If exposure time is short are we still not facing a constant exposure of light to acquire the correct exposure for the image? So adding to a short exposure to build up to a normal constant should give us a normal print right?
But what your saying is that due to the short exposure times I will run into noise anyway? In film, we used to build up to the normal exposure constant.
Aperture + Shutter Speed = C so if I was to take two separate photos for a multi do I halve this or not if I wanted to take lets say two shots of the same room and have a subject lets say a model for a portrait in two separate positions in the shot???
Then since I am still new to Digital Photography how do you figure your exposure equation to get the normal constant of light without the build up of noise? Now I am not even thinking of Painting with light yet, just multiple exposures.
Now to add to the confusion in my head. When I used to shoot with a dark background and do multiple I used to shoot the subject in the dark area. My second shot would also be in the dark area but I would expose both of these as normal and then as a result I would get two normal exposed images (next to each other) on the same photo. Can we still do this in Digital without noise buildup?
Not4wood
I am trying to understand this concept.
Going away from Reciprocity and turning this to noise control in multiple exposures. Or if you feel I should I could start a new thread about this??
So if longer exposure times will add to the noise, then multiple exposure should also add to it due to the build up of detail? Even before I start to experiment with multiple exposures I would like to be able to understand what I'm supposed to be facing here. If exposure time is short are we still not facing a constant exposure of light to acquire the correct exposure for the image? So adding to a short exposure to build up to a normal constant should give us a normal print right?
But what your saying is that due to the short exposure times I will run into noise anyway? In film, we used to build up to the normal exposure constant.
Aperture + Shutter Speed = C so if I was to take two separate photos for a multi do I halve this or not if I wanted to take lets say two shots of the same room and have a subject lets say a model for a portrait in two separate positions in the shot???
Then since I am still new to Digital Photography how do you figure your exposure equation to get the normal constant of light without the build up of noise? Now I am not even thinking of Painting with light yet, just multiple exposures.
Now to add to the confusion in my head. When I used to shoot with a dark background and do multiple I used to shoot the subject in the dark area. My second shot would also be in the dark area but I would expose both of these as normal and then as a result I would get two normal exposed images (next to each other) on the same photo. Can we still do this in Digital without noise buildup?
Not4wood
"tacit" wrote in message
In article <tJRTi.346$>,
"Luis Ortega" wrote:
I think that I understand the difference between noise and reciprocity failure.Longer exposure times will add to the noise; you'll see increased effects from thermal noise and from differences in sensitivity within the individual sensing elements of the CCD.
If you get digital noise when photographing in low light, will increasing exposure time reduce the noise or just add to it?
--
Photography, kink, polyamory, shareware, and more: all at http://www.xeromag.com/franklin.html
Not4wood wrote:
I am trying to understand this concept.
Going away from Reciprocity and turning this to noise control in multiple exposures. Or if you feel I should I could start a new thread about this??
So if longer exposure times will add to the noise, then multiple exposure should also add to it due to the build up of detail? Even before I start to experiment with multiple exposures I would like to be able to understand what I'm supposed to be facing here. If exposure time is short are we still not facing a constant exposure of light to acquire the correct exposure for the image? So adding to a short exposure to build up to a normal constant should give us a normal print right?
The thermal noise associated with the exposure builds up linearly with time, as does the image signal, and it is largest noise for low-light exposures. Thus the ratio of signal to noise (S/N)is constant. When multiple short exposures are combined, the signal and the noise increase by the root mean square, and the S/N improves as the square root of the number of images. Thus, combining 4 images improves S/N by a factor of 2, and combining 9 improves S/N by a factor of 3. The result is noise reduction, not noise removal.
But what your saying is that due to the short exposure times I will run into noise anyway? In film, we used to build up to the normal exposure constant.
Aperture + Shutter Speed = C so if I was to take two separate photos for a multi do I halve this or not if I wanted to take lets say two shots of the same room and have a subject lets say a model for a portrait in two separate positions in the shot???
I have done a lot of emission spectroscopy in which light intensities are measured by scanning photographs. S/N is fairly constant over most of the range of intensities, but increases at both the low and high ends. But when we are dealing with photographs, the way that the eye-brain senses light variations comes into significance, and the answer to your question is more complex. It is also somewhat subjective. Also, the granularity of the image is called noise.
Then since I am still new to Digital Photography how do you figure your exposure equation to get the normal constant of light without the build up of noise? Now I am not even thinking of Painting with light yet, just multiple exposures.
You figure exposures the same way. Many digicams let you automatically take sets of pictures that bracket the exposure, and image editing programs let you combine different parts of the image into one composite, akin to dodging in making a print photographically. Also, you can adjust the contrast in an image editing program.
Now to add to the confusion in my head. When I used to shoot with a dark background and do multiple I used to shoot the subject in the dark area. My second shot would also be in the dark area but I would expose both of these as normal and then as a result I would get two normal exposed images (next to each other) on the same photo. Can we still do this in Digital without noise buildup?
Not4wood
"tacit" wrote in message
In article <tJRTi.346$>,
"Luis Ortega" wrote:
I think that I understand the difference between noise and reciprocity failure.Longer exposure times will add to the noise; you'll see increased effects from thermal noise and from differences in sensitivity within the individual sensing elements of the CCD.
If you get digital noise when photographing in low light, will increasing
exposure time reduce the noise or just add to it?
--
Photography, kink, polyamory, shareware, and more: all at http://www.xeromag.com/franklin.html
Thanks Marvin.
So to just make sure. I shoot it the way we all used to and play so I can get the results I want. No sense over thinking this and blocking the actual photo work. Just to keep it in the back of my head.
Right? LOL
Mark
Not4wood
"Marvin" wrote in message
Not4wood wrote:
I am trying to understand this concept.
Going away from Reciprocity and turning this to noise control in multiple exposures. Or if you feel I should I could start a new thread about this??
So if longer exposure times will add to the noise, then multiple exposure should also add to it due to the build up of detail? Even before I start to experiment with multiple exposures I would like to be able to understand what I'm supposed to be facing here. If exposure time is short are we still not facing a constant exposure of light to acquire the correct exposure for the image? So adding to a short exposure to build up to a normal constant should give us a normal print right?
The thermal noise associated with the exposure builds up linearly with time, as does the image signal, and it is largest noise for low-light exposures. Thus the ratio of signal to noise (S/N)is constant. When multiple short exposures are combined, the signal and the noise increase by the root mean square, and the S/N improves as the square root of the number of images. Thus, combining 4 images improves S/N by a factor of 2, and combining 9 improves S/N by a factor of 3. The result is noise reduction, not noise removal.
But what your saying is that due to the short exposure times I will run into noise anyway? In film, we used to build up to the normal exposure constant.
Aperture + Shutter Speed = C so if I was to take two separate photos for a multi do I halve this or not if I wanted to take lets say two shots of the same room and have a subject lets say a model for a portrait in two separate positions in the shot???
I have done a lot of emission spectroscopy in which light intensities are measured by scanning photographs. S/N is fairly constant over most of the range of intensities, but increases at both the low and high ends. But when we are dealing with photographs, the way that the eye-brain senses light variations comes into significance, and the answer to your question is more complex. It is also somewhat subjective. Also, the granularity of the image is called noise.
Then since I am still new to Digital Photography how do you figure your exposure equation to get the normal constant of light without the build up of noise? Now I am not even thinking of Painting with light yet, just multiple exposures.
You figure exposures the same way. Many digicams let you automatically take sets of pictures that bracket the exposure, and image editing programs let you combine different parts of the image into one composite, akin to dodging in making a print photographically. Also, you can adjust the contrast in an image editing program.
Now to add to the confusion in my head. When I used to shoot with a dark background and do multiple I used to shoot the subject in the dark area. My second shot would also be in the dark area but I would expose both of these as normal and then as a result I would get two normal exposed images (next to each other) on the same photo. Can we still do this in Digital without noise buildup?
Not4wood
"tacit" wrote in message
In article <tJRTi.346$>,
"Luis Ortega" wrote:
I think that I understand the difference between noise and reciprocity failure.Longer exposure times will add to the noise; you'll see increased effects from thermal noise and from differences in sensitivity within the individual sensing elements of the CCD.
If you get digital noise when photographing in low light, will increasing
exposure time reduce the noise or just add to it?
--
Photography, kink, polyamory, shareware, and more: all at http://www.xeromag.com/franklin.html
Thanks Marvin.<snip>
So to just make sure. I shoot it the way we all used to and play so I can get the results I want. No sense over thinking this and blocking the actual photo work. Just to keep it in the back of my head.
Right? LOL
Mark
Not4wood
The multiple exposures approach is not going to work with moving subjects like traffic trails at night but might work with still objects. For light painting and light tracing, I find that the photo is already weird enough that the noise from a 1 to 5 minute exposure using the camera's lowest ISO setting is not an issue.
One field that uses multiple exposures is astrophotography and they do need to keep noise down so it might be useful for you to go ask questions at such user groups.
My curiosity was just academic since I approach it more from the art side of things and the noise I have seen in long exposure images shot at the lowest ISO setting are not a problem for me.<snip>
There is one thing you can do with moving subjects. Repeat the exposure with the lens cap on, which gives you the dark noise pixel-by-pixel. Subtracting the dark field image from the real one improves S/N by 40%.
The multiple exposures approach is not going to work with moving subjects like traffic trails at night but might work with still objects. For light painting and light tracing, I find that the photo is already weird enough that the noise from a 1 to 5 minute exposure using the camera's lowest ISO setting is not an issue.You could always try the technique that astronomers use to reduce thermal noise. Chill the sensor in liquid nitrogen. Liquid helium is harder to come by.
One field that uses multiple exposures is astrophotography and they do need to keep noise down so it might be useful for you to go ask questions at such user groups.
My curiosity was just academic since I approach it more from the art side of things and the noise I have seen in long exposure images shot at the lowest ISO setting are not a problem for me.
On Sun, 28 Oct 2007 09:32:53 GMT, "Luis Ortega" wrote:I can see ways to build a digicam to do that, but I can't see how to do it with any off-the-shelf camera. The sensor could be chilled with a thermoelectric cooler, if you can plug the camera in to a big enough power source.
The multiple exposures approach is not going to work with moving subjects like traffic trails at night but might work with still objects. For light painting and light tracing, I find that the photo is already weird enough that the noise from a 1 to 5 minute exposure using the camera's lowest ISO setting is not an issue.You could always try the technique that astronomers use to reduce thermal noise. Chill the sensor in liquid nitrogen. Liquid helium is harder to come by.
One field that uses multiple exposures is astrophotography and they do need to keep noise down so it might be useful for you to go ask questions at such user groups.
My curiosity was just academic since I approach it more from the art side of things and the noise I have seen in long exposure images shot at the lowest ISO setting are not a problem for me.
John
There is one thing you can do with moving subjects. Repeat the exposure with the lens cap on, which gives you the dark noise pixel-by-pixel. Subtracting the dark field image from the real one improves S/N by 40%.
They cool the detector to reduce dark noise. Not possible with a digicam.
"Marvin" wrote in message
There is one thing you can do with moving subjects. Repeat the exposure with the lens cap on, which gives you the dark noise pixel-by-pixel. Subtracting the dark field image from the real one improves S/N by 40%.
Thanks, that's an interesting idea and I'll give it a try. Would you please explain how to subtract the dark field image from the exposed image?
Do you place them both on the same file in different layers and use some blending mode in Photoshop?
Battery voltage would decrease faster than the noise. And the noise reduction wouldn't be much.They cool the detector to reduce dark noise. Not possible with a digicam.
I would think then that shooting outdoors in winter will help to reduce the noise on an image even if just a little?
I suppose that it can affect battery life too.
Are digital camera exposures affected by reciprocity failure the way that film exposures are?
Thanks for any advice.
On Oct 24, 5:15 am, "Luis Ortega" wrote:
Are digital camera exposures affected by reciprocity failure the way that film exposures are?
Thanks for any advice.
Yeesh there are a lot of trolls on this board. i see sandcastle kickers and piddlers on the post toastsie. weeee
Ok sorry had to get that out. lol
No, DSLR sensors do not have reciprocity failure the way film does. Reciprocity failure occurs because of a chemical change on the emulsion layer of film at long exposures.
The sensors have their own set of issues, different than reciprocity failure.
On Oct 24, 5:15 am, "Luis Ortega" wrote:
Are digital camera exposures affected by reciprocity failure the way that film exposures are?
Thanks for any advice.
Yeesh there are a lot of trolls on this board. i see sandcastle kickers and piddlers on the post toastsie. weeee
Ok sorry had to get that out. lol
No, DSLR sensors do not have reciprocity failure the way film does. Reciprocity failure occurs because of a chemical change on the emulsion layer of film at long exposures.
The sensors have their own set of issues, different than reciprocity failure.
Noise occurs in areas of underexposure at all ISOs and to varying degrees at higher ISOs in the different camera models/sensors. Even within systems that varies. Example: Rebel XT vs a 5D. My XT at ISOs under 400 handles exposures beautifully as LONG AS the exposure is slam on. Underexposure means noise, pure and simple. Above 400 its even more critical, and the noise is starting to be noticeable at all the higher ISOs. Noiseware is a must in editing.
My 5D, noise isn't barely present at any ISO, even up to 1600. Only with severe underexposure in the shadows does it become noticeable, and then still not near at the level of the XT.
the 10d, 20d, 30d and 40d, handle noise better than the XT but not as well as the 5D *from all the images I have seen coming out of the cameras*. I suspect the 40d is closer to the 5d in its handling than its predecessors.
Solution? Expose digital like chrome: expose for the highlights and develop for the shadows, realizing that in "general" digital's dynamic range is 5-6 stops.Realistically staying within 4 stops will keep details in both the highlights and shadows, and shooting in RAW will allow for keeping details in both where tweaking is needed. In situations where you are dealing with landscapes you will need to deal with the images as HDR images and expose multiple times to keep all the detail throughout the scene and merge them in post processing.
Hope that helps some and isn't too verbose.
Kathie
Art Beat Photography
Thanks.
In long exposures at night, where a lot of areas will be underexposed except for the traffic trails or light sources, is there any suggested procedure to minimize noise in the dark areas?
I tend to just go with my lowest ISO setting and use similar times to when I use film.
Rob wrote:
Luis Ortega wrote:
Are digital camera exposures affected by reciprocity failure the way that film exposures are?
In article <AHJTi.13341$>,
Marvin wrote:
Rob wrote:
Luis Ortega wrote:
Are digital camera exposures affected by reciprocity failure the way that film exposures are?
*
Read this:
http://silvergrain.org/Photo-Tech/reciprocity.html
earle
*
I doubt any consumer DSLR is any good for exposures past a few seconds.