Velocity Reviews - Computer Hardware Reviews

Velocity Reviews > Newsgroups > Computing > Digital Photography > Reciprocity failure in digital camers?

Reply
Thread Tools

Reciprocity failure in digital camers?

 
 
g@risky-biz.com
Guest
Posts: n/a
 
      09-12-2006
I like to take night shots. I've had some success at it with my film
SLRs, but there was always some voodoo involved; taking several shots
on B, holding the shutter open for 2 seconds, 5 seconds, 10 seconds,
etc.

I just got a Kodak P880. I chose it mostly for the wide angle
capability and the reasonable price. While on vacation in Montreal and
Quebec City I got to experiment a bit with it. I have to say that the
ability to view the night shots on the screen was a real blessing. It
allowed me to fine-tune the exposure and be sure I got a decent shot.

I have to say that sometimes (not all the time) it was almost too easy.
In certain modes the auto-exposure will use shutter speeds up to 16
seconds. It got the exposure pretty close to "right" a fair percentage
of the time.

That notwithstanding, I'm curious about reciprocity with digital
sensors, if only to know what is going on a bit better. Is there a
point at which doubling the exposure time doesn't produce a "one stop"
change, as with film? Or makes a color change?

I also did a litle experimenting with the ISO settings but not in a
systematic way, I'm afraid. Assuming no camera shake, and a stationary
subject (buildings, etc.), would you use a lower ISO and longer
exposure or the other way around? I'm wondering which results in less
noise.

Greg Guarino

 
Reply With Quote
 
 
 
 
acl
Guest
Posts: n/a
 
      09-12-2006
http://www.velocityreviews.com/forums/(E-Mail Removed) wrote:
> I like to take night shots. I've had some success at it with my film
> SLRs, but there was always some voodoo involved; taking several shots
> on B, holding the shutter open for 2 seconds, 5 seconds, 10 seconds,
> etc.
>
> I just got a Kodak P880. I chose it mostly for the wide angle
> capability and the reasonable price. While on vacation in Montreal and
> Quebec City I got to experiment a bit with it. I have to say that the
> ability to view the night shots on the screen was a real blessing. It
> allowed me to fine-tune the exposure and be sure I got a decent shot.
>
> I have to say that sometimes (not all the time) it was almost too easy.
> In certain modes the auto-exposure will use shutter speeds up to 16
> seconds. It got the exposure pretty close to "right" a fair percentage
> of the time.
>
> That notwithstanding, I'm curious about reciprocity with digital
> sensors, if only to know what is going on a bit better. Is there a
> point at which doubling the exposure time doesn't produce a "one stop"
> change, as with film? Or makes a color change?


No, there is no colour shift or loss of sensitivity. There is an
increase of hot pixels; some cameras have a dark frame subtraction
feature which eliminates them (by taking another exposure of the same
length with the shutter closed immediately afterwards and subtracting
that). This usually doubles the exposure time (since the second exposure
is of the same length). Of course, you can simply not use this (I don't
with my D200 for ISOs below 400 and shutter speeds up to about half a
minute). Experiment!

>
> I also did a litle experimenting with the ISO settings but not in a
> systematic way, I'm afraid. Assuming no camera shake, and a stationary
> subject (buildings, etc.), would you use a lower ISO and longer
> exposure or the other way around? I'm wondering which results in less
> noise.


As for this, I imagine it depends on the camera. Probably best to
experiment with yours and decide based on that.
 
Reply With Quote
 
 
 
 
Matt Ion
Guest
Posts: n/a
 
      09-12-2006
(E-Mail Removed) wrote:

> That notwithstanding, I'm curious about reciprocity with digital
> sensors, if only to know what is going on a bit better. Is there a
> point at which doubling the exposure time doesn't produce a "one stop"
> change, as with film? Or makes a color change?


Well, here are a couple examples:

http://www.pbase.com/image/34152731

"This is a long (almost 9 minutes) exposure that I took last night using only
the light from the moon." - Annika1980 (circa Sept 2004)

Here's one I took under moonlight as well:
http://www.photosig.com/go/photos/vi...KI2?id=1716559

> I also did a litle experimenting with the ISO settings but not in a
> systematic way, I'm afraid. Assuming no camera shake, and a stationary
> subject (buildings, etc.), would you use a lower ISO and longer
> exposure or the other way around? I'm wondering which results in less
> noise.


Lower ISO will generally mean less noise, but remember it also needs a wider
aperture for the same exposure time, which will lose some DOF, and that a longer
exposure will eat up more batteries.
 
Reply With Quote
 
Marvin
Guest
Posts: n/a
 
      09-12-2006
(E-Mail Removed) wrote:
<snip>
> That notwithstanding, I'm curious about reciprocity with digital
> sensors, if only to know what is going on a bit better. Is there a
> point at which doubling the exposure time doesn't produce a "one stop"
> change, as with film? Or makes a color change?

<snip>
> Greg Guarino
>


The silver halide particles in film must absorb 2 photons
within a short time interval for the particle to become a
silver metal particle in development. That is the cause of
reciprocity failure, and it does not happen with a digital
sensor. When a sensor pixel has reached the point where it
can't hold more charge, it is said to be saturated. Near
there, doubling the exposure time will produce less of a
change. Saturation also occurs with film. It is not
reciprocity failure.
 
Reply With Quote
 
Stormlady
Guest
Posts: n/a
 
      09-12-2006

<(E-Mail Removed)> wrote in message
news:(E-Mail Removed) ups.com...
>
> I also did a litle experimenting with the ISO settings but not in a
> systematic way, I'm afraid. Assuming no camera shake, and a stationary
> subject (buildings, etc.), would you use a lower ISO and longer
> exposure or the other way around? I'm wondering which results in less
> noise.
>
> Greg Guarino
>


I have the Kodak P850 (for the zoom) and I have found that in certain shots,
going over iso 200 produced a lot of noise, really an unreasonable amount.
For example, we were at the fair with our daughter and she was on a ride,
Trying to get the shot of her moving, boyfriend turned the iso up to 200 and
the pictures are really pretty bad. Unusably bad because of the noise.
That being said, I've taken pictures of other things at the higher iso and
not gotten them so noisy, things that weren't moving as much.


 
Reply With Quote
 
Ron Baird
Guest
Posts: n/a
 
      09-19-2006
Greetings Greg,

Glad to hear about the shots of Montreal. Great city and the P880 could
capture most all scenes.

You do not have to worry about reciprocity in digital, variations are
accounted for in the camera. Reciprocity is really more related to film and
some paper. With Film, if you use long exposures the sensitivity of the film
is reduced slightly which can cause a shift in color. With a digital camera
that problem is removed and most cameras can handle exposures of several
seconds or longer.

What you should consider is that low light or higher ISO settings in digital
camera can yield an increase in 'noise' the digital equivalent of grain. You
may want to get a good filter for it, or in your case improve on some shots
with flash. The P20 can talk to your camera and would be an advantage.

Talk to you soon,

Ron Baird
Eastman Kodak Company




<(E-Mail Removed)> wrote in message
news:(E-Mail Removed) ups.com...
>I like to take night shots. I've had some success at it with my film
> SLRs, but there was always some voodoo involved; taking several shots
> on B, holding the shutter open for 2 seconds, 5 seconds, 10 seconds,
> etc.
>
> I just got a Kodak P880. I chose it mostly for the wide angle
> capability and the reasonable price. While on vacation in Montreal and
> Quebec City I got to experiment a bit with it. I have to say that the
> ability to view the night shots on the screen was a real blessing. It
> allowed me to fine-tune the exposure and be sure I got a decent shot.
>
> I have to say that sometimes (not all the time) it was almost too easy.
> In certain modes the auto-exposure will use shutter speeds up to 16
> seconds. It got the exposure pretty close to "right" a fair percentage
> of the time.
>
> That notwithstanding, I'm curious about reciprocity with digital
> sensors, if only to know what is going on a bit better. Is there a
> point at which doubling the exposure time doesn't produce a "one stop"
> change, as with film? Or makes a color change?
>
> I also did a litle experimenting with the ISO settings but not in a
> systematic way, I'm afraid. Assuming no camera shake, and a stationary
> subject (buildings, etc.), would you use a lower ISO and longer
> exposure or the other way around? I'm wondering which results in less
> noise.
>
> Greg Guarino
>



 
Reply With Quote
 
Alan Meyer
Guest
Posts: n/a
 
      09-19-2006
Ron Baird wrote:
> ...
> What you should consider is that low light or higher ISO settings in digital
> camera can yield an increase in 'noise' the digital equivalent of grain. You
> may want to get a good filter for it, or in your case improve on some shots
> with flash. The P20 can talk to your camera and would be an advantage.
> ...


I'm not an expert on this, but I hypothesize that the explanation
for Ron's comment is as follows:

A digital sensor accumulates charge (i.e., electrons) in each
pixel position as a response to light falling on the sensor in that
position.

The sensor also accumulates charge in each pixel simply as
a result of the random motion of electrons (due to heat,
radiation, and perhaps other factors other than stimulation
by light.)

We see digital "noise" when the ratio of random charge becomes
a significant percentage of the total charge in a pixel, causing
the value of that pixel to have a noticeable random component
to it.

The sensor accumulates random charge as a function of time.
The sensor is discharged just before the image is made, and
read after the image is made.

In bright light, the time between pre-image discharge and
post-image reading is short, allowing little random charge
to accumulate.

In dim light the time is long, allowing more random charge
to accumulate. Hence the extra noise in low light situations.

Dark areas of an image are most subject to this because
the random charge in all areas of the image is the same, but the
dark areas have a higher percentage of random to light
stimulated charge accumulation.

The "filtering" that Ron talks about is the digital process of
averaging adjacent pixels with similar values in order to
eliminate differences that are due to random factors. But
no filter, no matter how smart, can truly distinguish
differences due to random charge as distinct from differences
due to different amounts of light. It uses heuristic algorithms
to make intelligent guesses, which are often but not always
right. Hence there is always some loss of detail when
aggressive noise filtering is performed.

Alan

 
Reply With Quote
 
 
 
Reply

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
I don't *think* this is reciprocity failure... Paul Ciszek Digital Photography 7 09-16-2012 01:49 AM
time.strptime() and time.strftime() reciprocity Greg Krohn Python 1 02-17-2009 12:26 AM
Reciprocity a problem with digital? jimphotog via PhotoKB.com Digital Photography 4 03-22-2007 11:02 AM
[JAVA] [EVALUATION] - The Java Failure (Sorry: The Java(tm) Failure) Ilias Lazaridis Java 0 02-01-2005 10:32 AM
reciprocity failure - what is it? cc Digital Photography 27 01-18-2004 09:42 PM



Advertisments