Time of exposition versus ISO value

[Panama Hat & Camera]

What is the best to have less noise and to increase image quality (maintaining constant the other variables and using a sturdy tripod): increase the time of exposition or increase the ISO value?
For example:
f = 300m - f/5.6 - 1/16 s - ISO 1600 or
f = 300m - f/5.6 - 1 s - ISO 100 ?
Or the choice depends upon the brand and model of the camera?
Best regards,
Antonio.

[Mike Buckley]

I also used to wonder about that. Using my camera bodies, I regularly use properly exposed images of up to 30 seconds and the base ISO with absolutely no noise. I might be able to use even longer exposures but have never had the need to try.

[Shadowman]

What is the best to have less noise and to increase image quality (maintaining constant the other variables and using a sturdy tripod): increase the time of exposition or increase the ISO value?
For example:
f = 300m - f/5.6 - 1/16 s - ISO 1600 or
f = 300m - f/5.6 - 1 s - ISO 100 ?
Or the choice depends upon the brand and model of the camera?
Best regards,
Antonio.

With the settings you've chosen and your particular model of camera, the difference will mean a bit more processing (noise reduction, contrast) if shot at higher ISO, this also depends on the particular subject and time of day, and whether or not there is some subject movement.

[DanK]

Unless there is a particular reason to use a higher ISO (for example, if motion requires a shorter exposure), a lower ISO is better. Increasing ISO reduces dynamic range and increases noise.

Of course, a longer exposure means that you have to control for motion. Sometimes a long exposure is a problem even with a tripod, for example, in a strong wind. However, when one can control motion, long exposures are fine. The picture I will post below was a 425 second exposure near midnight, under a full moon.



I routinely

[dem]

I don't think there is a camera that would produce a better image at ISO 1600 and 1/16s than at ISO 100 and 1s. What factors would contribute to the degradation of the image at ISO 100 more than at ISO 1600? Any correlated noise (e.g. hot pixels) will be present in a similar amount but uncorrelated noise (e.g. chroma) will be lower at ISO 100. If subject movement and camera shake are not a problem why choose higher ISO?

[DanK]

I don't think there is a camera that would produce a better image at ISO 1600 and 1/16s than at ISO 100 and 1s. What factors would contribute to the degradation of the image at ISO 100 more than at ISO 1600? Any correlated noise (e.g. hot pixels) will be present in a similar amount but uncorrelated noise (e.g. chroma) will be lower at ISO 100. If subject movement and camera shake are not a problem why choose higher ISO?

and in addition, increasing ISO results in smaller dynamic range.

Put differently: decreasing exposure and increasing ISO means decreasing signal and using amplification to compensate. No way that can improve things, assuming stable camera and subject.

[pnodrog]

If the camera is absolutely still, the scene is static and lighting is constant, the closer to the camera's native ISO you use the better the dynamic range will be and the lower the noise should be. Usually other factors make the choice a little more complicated.

[Manfred M]

Antonio - I can only re-iterated what the others have said so clearly. Shooting at base ISO (or as close to it as possible) will give you the best image quality, with low noise, high dynamic range and high colour depth being the chief advantages.

That being said, this also assumes you have a way of holding the camera steady and the subject is reasonably static and will not cause you other problems by moving during the exposure.

I would suggest these are "universal" answers and are independent of make and model of camera.

[Loose Canon]

Doesn't his depend on what you are shooting and under what kind of lighting Antonio?

Even my Canon 50D shoots well up to 1600 ISO with the right exposure settings and its a fairly old model by recent standards.

[Loose Canon]

This is the second time I have seen the term "exposition" on this forum. I had to look it up.

I'm not seeing any definition that relates to photography.

[Inkanyezi]

This is the second time I have seen the term "exposition" on this forum. I had to look it up.

I'm not seeing any definition that relates to photography.

That, of course, is a purely linguistic curiosity. As we all know, the world is full of different languages, and in some of them, similar expressions are worded somewhat differently. In Brazil, Portuguese is spoken, and exposure is exposição, which is not far from exposition, which of course is not unknown in photography, even though it has a different meaning than exposure. (In Spanish the term for exposure is exposición.)

So, for better learning of English terminology and the English language, it is worth pointing out, but the slight error does not impede understanding the question.

I think most of the answers respond well to the OP question, with the caveat that dynamic range perhaps could need more explanation. What is it, when would I need more of it and when can I jettison dynamic range, to gain other desired properties? And what would make noise a problem worth pondering?

[DanK]

it is worth pointing out, but the slight error does not impede understanding the question.

I agree. It appears that everyone who posted understood what Antonio meant. To use an American idiom, "let's cut them some slack." It's hard to post in a second (or third or fourth) language without errors. I certainly would not be able to do it in either of the languages I learned.

Doesn't his depend on what you are shooting and under what kind of lighting Antonio?

I think you are answering a different question than the one Antonio posed. Antonio did not ask whether a higher ISO might get him results similar enough for some practical purposes. I think that is the question you are answering. He asked,

What is the best to have less noise and to increase image quality

Assuming no camera or subject motion, the combination of higher ISO and lower shutter speed cannot improve image quality. There may be some circumstances where no degradation is apparent--for example, for an image with small tonal range, exposed to the right, and displayed small. I and others have shown this with series of photos posted here. However, that's a different question.

[Loose Canon]

It appears that everyone who posted understood what Antonio meant. To use an American idiom, "let's cut them some slack."

I think you are answering a different question than the one Antonio posed.

Well Dan, it appears that not everybody understood the question if you think I didn’t. And I probably don’t which is why I had to look up some verbiage used to see what I was missing, if anything.

I probably still don't understand and using a term that I couldn't find a meaning that relates didn't help.

So while you are cutting slack you might cut me a little as well.

Antonio I apologize if I cause you any offense. That was certainly not my intention.

[DanK]

Terry,

I wasn't meaning to criticize you--just expressing that we (all) need to recognize that posting in English isn't easy for some of our participants. Sorry if I came across differently. I actually had rewritten the post to try not to sound critical. Obviously, I didn't succeed. Sorry.

My second point is really unrelated. You can substitute the correct "exposure" into Antonio's post, and it wouldn't change the point. I think there are two different issues to consider--which ISO is better, and does it always make a difference--and it is worth keeping them separate. I have posted several times making the point you made, which is that it often makes little or no difference which ISO you choose, within limits. But it is a different question.

So again, I apologize that I offended you.

Dan

[xpatUSA]

That, of course, is a purely linguistic curiosity. As we all know, the world is full of different languages, and in some of them, similar expressions are worded somewhat differently. In Brazil, Portuguese is spoken, and exposure is exposição, which is not far from exposition, which of course is not unknown in photography, even though it has a different meaning than exposure. (In Spanish the term for exposure is exposición.)

So, for better learning of English terminology and the English language, it is worth pointing out, but the slight error does not impede understanding the question.

I think most of the answers respond well to the OP question, with the caveat that dynamic range perhaps could need more explanation. What is it, when would I need more of it and when can I jettison dynamic range, to gain other desired properties? And what would make noise a problem worth pondering?

Well said, Urban.

I find the phrase "dynamic range" especially irritating when it is stated without qualification, leaving the reader to guess which of it's several definitions is applicable.

[DanK]

I find the phrase "dynamic range" especially irritating when it is stated without qualification, leaving the reader to guess which of it's several definitions is applicable.

Uh-oh. I do that. I use it to mean the range of luminance that an image contains or that a sensor can capture. Can you explain the different meanings?

[Mike Buckley]

Add to Dan's two possibilities the dynamic range of the scene being captured.

[xpatUSA]

Uh-oh. I do that. I use it to mean the range of luminance that an image contains or that a sensor can capture. Can you explain the different meanings?

Would take too long Dan and would likely open a can worms.

Some examples starting with yours above:

Image content is easy: the ratio (not actually range) of the lightest to the darkest tone in the image. Except for the fact that 255/0 is infinity so the darkest tone can not be zero.

And 66535/0 for a 16-bit image is an even bigger infinity

In which case, what is a reasonable "darkest tone"? Is it 1? Is it -9 EV?

Or is the dynamic range of an image that which comes out of the screen as we edit away? In which case, what if the DR of the screen itself is less than the tonal range of image being displayed? For example many monitors promise a contrast of 1000:1 (white cd/m2 over black cd/m2 = 60dB or ~10EV). But many folks, myself included, turn down the brightness and (NEC) turn up the black level a bit, bringing that contrast (and therefore perceived dynamic range) down quite a lot. I digress.

Or, is the said image a print, viewed under unknown lighting in which case the "dynamic range" is anything you like

"or that a sensor can capture":

That could refer to a scene dynamic range or it could refer to the capability of the sensor itself.

'Scene' is perhaps the easiest of all DR definitions: brightest part of the scene to the darkest. Potentially a pretty big number as we all should know (no wonder dB are the popular units for that) . . . sun versus black cat in a coalmine for example/

'Sensor capability': here we go again. Are we talking just about the sensor? Are we including noise from all sources? What upper limit do we place?: is that limit the simple point of saturation (well full) or some margin below it? What margin?: is it the exact point at which the signal goes non-linear, or something based on photon (shot) noise? Or do we lump in all the camera electronics, including 'read noise', with all of the foregoing?

By answering your question with a flurry of rhetorical questions, all having answers that most of us know, I am not deliberately being a smart-ass. I'm just trying to show that an unqualified "DR" tells us little unless we are comparing like with like - such a camera tests on DPR or similar. In a series of posts from various authors within in a thread, "like with like" is quite unlikely

Having said all that, there is an ISO definition for the DR of a camera, for what it's worth. The well-respected Doug Kerr has a lot to say about it, starting with:

"The dynamic range of a digital camera can be simplistically defined as the ratio of the maximum and minimum luminance that a camera can “capture” in a single exposure. But, when we try to quantify this property, we find that the establishment of an explicit definition is much more complicated than it seems on the surface. International Standard ISO 15739-2003 gives an explicit definition of dynamic range for a digital still camera and a procedure for determining it"

http://dougkerr.net/Pumpkin/articles...amic_range.pdf

Worth a read. Amongst other things, we learn that the ever-popular low bound for DR (signal is equal to noise) is impossible to measure with a sensor.

[DanK]

Ted,

Thanks. From a quick scan, I think this is a can of worms I will enjoy.

Just glancing at it--have to get back on task now--I have a few quick thoughts.

1. In my work, there is always a difference between the construct one intends to measure and the operationalized measure of it. In some areas, these two can be treated as the same, but not in my work. I think your point could be translated into my terminology this way: the definition is clear with respect to the construct, but the actual measurement is not, for any number of reasons, some of which have to do with particular operationalizations.

2. if so, then this suggests that is is not the concept of DR that is hazy, but rather it's concrete application.

3. I think it is a range when, as usual, it is expressed in stops, which is a log scale. The log of a ratio is a difference.

Anyway, that's just off the top of my head, so it may be way off track. I will mull this over when I have more time to read what you sent.

Dan

[xpatUSA]

Ted,

I have a few quick thoughts.

1. In my work, there is always a difference between the construct one intends to measure and the operationalized measure of it. In some areas, these two can be treated as the same, but not in my work. I think your point could be translated into my terminology this way: the definition is clear with respect to the construct, but the actual measurement is not, for any number of reasons, some of which have to do with particular operationalizations.

2. if so, then this suggests that is is not the concept of DR that is hazy, but rather it's concrete application.

Agreed. The discussion occurs quite often on DPR's Sigma forum where it often happens that people like me are talking about X3F raw data (no NR involved) but others are talking about what comes out of Sigma's proprietary converter (which has really good NR and needs it).

I think it is a "range" when, as usual, it is expressed in stops, which is a log scale. The log of a ratio is a difference.

Indeed - I confess that the word "range" ruffled my feathers for a long, time. Only recently did I realize the same thing, duh.