Is the goal to have the lowest ISO possible

[apaflo]

but given all things being equal, we should adjust the ISO last, since it essence, it degrades the picture.

Well... all things aren't always equal though. Sometimes we set the ISO first, knowing very well we need to start with a limit on how much degradation we will accept as a result of too low an exposure. Then we set exposure even though it might not give us precisely the artistic effects desired.

For example, event photographers commonly shoot in low light. At one event there might be little motion, so a shutter speed of 1/100 could be sufficient, and that may also allow stoppig down for more depth of field. The first parameter set would have been the highest acceptable ISO. And if something then gets going where it is more important to freeze action, the ISO value is retained, but the shutter speed must be set faster. Rather than increase ISO (a disaster), the aperture is set wider even though it is detrimental. In these cases ISO is set first, it is the make or break paramenter, and it doesn't get changed when the environment changes.

Compare that to shooting outdoor events with a camera such as the Nikon D800, where the dynamic range is high enough that changing from ISO 200 to 400 or even to 800 simply will not affect the image that is saved in a JPEG image format which can only encode about 9 f/stops of dynamic range. In this case ISO would be the first thing to change, rather than exposure. An ISO change has no effect, an exposure change is detrimental.

You are right that there are "too many variables", but as you gain experience and knowledge some of what seems overwhelming now will become "the basics" that you don't blink an eye at. Everyone at any given time has a limit. And you'll notice that those with a lower limit often claim anything above theirs is useless trivia. My suggestion is to make an effort, for the rest of your life, at raising your limit! That's what makes photograph fun.

 

[apaflo]

How soon do real ISO problems kick in? Obviously at the upper range there's significant noise: but on a modern body (sat a T2i), is there really much quality difference between ISO100, ISO200, and ISO400?

That is a very significant point. I can answer only for a few cameras, but it isn't too hard for someone to extrapolate from one to another or to actually go find the right data for a specific camera in order to know exactly.

Lets consider that the JPEG format can only encode about 9 stops of dynamic range, at most. And either a montior or a print can display about 8, at most. In almost all cases the actual useful dynamic range is a bit less than 8 stops. (If the image has more than that, it is almost always displayed with the excess range blocked off in the shadows.)

Because of that a camera with more than 8 stops of dynamic range has ISO to spare. One way to use the extra ISO is to allow exposure to be less critical. A camera with 12 stops has 4 to spare, and exposure can be really sloppy! A camera with 8 requires very precise exposure.

Therefore it is interesting to know just exactly at which ISO a camera drops to 8 stops of dynamic range. The Nikon D4 and the Nikon D800 and the Canon 1DX are within 1/4 of a stop of having 8 stops of dynamic range at ISO 1600. Virtually all other full frame cameras hit that mark at ISO 1000 (The exception is the Nikon D700 which is a hair lower than the others.).

Hence that is the ISO at which, with very precise exposure makes little difference. Going to 1 stop less ISO removes the extreme critical aspect of ISO (looking at the histogram on almost any shot be more accurate than that). And at 2 stops less almost any bit of slop is close enough

 

[Gavjenks]

In fact none of it is semantics. All of it is technically correct. All of it directly affects the image quality of capture data. And until a photographer does understand the significance of it (though not necessarily the technical details) their photography will be limited by that lack of understanding.

Yes, it is technically correct.

But No, nobody's photography will be limited no matter which way they see it. Thinking about ISO as "exposure at the cost of more noise" will lead to precisely the same decisions in the field as thinking of it as "a noise producing gain applied to a given exposure."


(By the way, here's an idea: if we simply put "exposure" in quotation marks, then everybody is happy, because as I learned the other day, doing so apparently instantly makes it mean whatever we want and renders it automatically unconfusing to any newbies!)

 

[fokker]

In a sense, yes, the goal is to use the lowest possible ISO - that's how I've always thought of it anyway. I will always try to use ISO100 whenever the lighting and subject determines that it is feasible. If my shutter speed is too low for my desired aperture then up goes the ISO until it is fast enough.

 

[apaflo]

But No, nobody's photography will be limited no matter which way they see it. Thinking about ISO as "exposure at the cost of more noise" will lead to precisely the same decisions in the field as thinking of it as "a noise producing gain applied to a given exposure."

Look into it a little more. Consider photon noise too. Things start making a lot of sense if and only if we keep exposure separate from ISO.

(By the way, here's an idea: if we simply put "exposure" in quotation marks, then everybody is happy, because as I learned the other day, doing so apparently instantly makes it mean whatever we want and renders it automatically unconfusing to any newbies!)

Very close! It renders it non-arbitrary to those who are astute. And in fact any time "exposure" is related to ISO it should be in quotes.

 

[bratkinson]

As someone mentioned on this forum several months ago, photography is an 'acceptable compromise'. ISO speed is merely one aspect of the photography triangle - aperture, shutter speed, and ISO speed.

Depending on the shooting conditions, say, a bright sunny day, there's perhaps 200 or more possible combinations of shutter speed, aperture and ISO speed that will produce an acceptably exposed photograph. I could, for example, decide to use 1/2000 shutter speed, f 32 aperture, and an appropriate ISO speed to yield the desired image brightness. Slowing the shutter speed by one stop and opening the aperture 1 stop will still yield the same results.

However, as mentioned by previous posts in this thread, there may be other considerations that will affect ones choices of settings. Getting a thin DOF, for example, starts with a wide aperture. Everything else has to be set with that aperture priority. Shooting in low light as I often do, my biggest concern is to freeze motion, so I start with a shutter speed of 125 or so. I'm willing to sacrifice ISO speed going high up to a limit, which varies by camera model. When I've reached that ISO limit, I can compensate by opening up the aperture. But I may not want a 2" DOF at f2.8 when zoomed out to 200mm, for example. Now what? I can slow down the shutter speed, risking motion blur. I can crank up the ISO speed getting more noise. Pick one. Pick both. That's the 'acceptable compromise'.

So, to state that using the lowest ISO speed is a goal would be like shooting with ASA 25 film in a modern world. Yes, ASA 25 gives fantastic color rendition. But by shooting with that slow an ASA, or even at ISO 100 in digital cameras, is strongly limits the kinds of shots that can be taken, or, requires lots and lots of light to be used successfully.

 

[jaomul]

If shooting a scene on a tripod such as a low light landscape, the native iso of 100 or 200 depending on model is normally better for dynamic range than low expanded 50 or whatever values are in low expanded. So lowest iso settings possible are not always the best choice

 

[JoeW28]

I realize given all the possibilities that the answer is no, but hypothetically, all other things being equal, should the ISO be the last thing I adjust to allow the light in. Maybe a better way to ask, should ISO be the last thing I go to to get the right exposure for the shot?

No--not as a hard and fast rule. If you're a photojournalist or sports shooter, you go with certain assumptions and you set to try allow you to shoot fast and adjust fast. But for the everyone else, you look at the setting and start to adjust accordingly. You want to emphasize motion? Than shutter speed will be the last thing you set....and you initially set other settings to allow you to control shutter speed (so you might open up the aperture and raise the ISO before setting shutter speed). You love the natural ambient light and it's a static setting? Than aperture might be the last thing you set after you've adjusted ISO and shutter speed and WB. You are shooting a picture with high contrast and you intend to blow it up (or crop extensively)? Then ISO will be the key for you (b/c you'll want to control for noise and pixelation/grain to allow for a blowup).

Also, I'm not so sure I'd say you set the most important thing last. If I know that a particular aspect is the key part of the shot for me (such as Dof) then I'll set aperture first and then monkey with everything else so I can keep the aperture setting I intend to use. So sometimes the key element (be it ISO, WB, DoF/aperture, shutter speed, filter, light source) will be what I set first...and then adjust around it to keep the first (and most critical factor) stable.

 

[DSRay]

No.

 

[imagemaker46]

In an ideal situation the lower the better, in a realistic world it doesn't always happen. If all you shoot is outdoors under great light, using a tripod and static objects yep. But what it really all comes down to, putting the cameras elements together to end up with a suitable situation to capture what you are shooting. Sometimes sacrificing the iso to get the speed and f stop is necessary. For example I know that 9 times out of 10 when I walk into an indoor sports venue the iso is going to be over 1000, just to get enough speed to stop the action.

It was like shooting film, using an asa film of 100-200 was great for bright sunlight, a mid range was asa 400-800 and then grain the size of golf ball dimples was 1600-3200. What ever was necessary.

 

[manaheim]

Here's a mind boggling thought... ISO has no direct effect on exposure.

Exposure is changed only by two things: one is how much light comes through the lens and the other is how long that light is allowed to hit the sensor. Basically aperture and shutter speed.

That results in a set amount of analog signal being generated by the electronic sensor. Nothing else changes that.

But after that analog signal is generated the camera does two things which may change the brightness of the recorded data. That doesn't really change the "exposure", but many people don't realize it actually is different. First, the signal might be amplified by an analog device, and second it is digitized with an Analog-Digital-Converter. In the process of being digitized, or afterwards, the digital data can be amplified too, just as the analog signal was. The significance to your original question is that the analog, the ADC amplification, and the post-ADC amplification each have different characteristics and all of them result in changes to brightness but otherwise lower quality for the image compared to what would have happened if the exposure had been increased. When done in the camera those are all called an increase in ISO (sensitivity to light, or how bright the recorded data is).

Whether that's right or wrong, it's mostly just semantics. It doesn't really alter the advice that is being provided. From a purely practical level, ISO has a negative effect on image quality. How that negative effect gets there mechanically is interesting, but not relevant to the question.

In fact none of it is semantics. All of it is technically correct. All of it directly affects the image quality of capture data. And until a photographer does understand the significance of it (though not necessarily the technical details) their photography will be limited by that lack of understanding.

You're making it overly complicated. We don't need to know the physics behind why the knife cuts, to know that it does.

 

[runnah]

Whether that's right or wrong, it's mostly just semantics. It doesn't really alter the advice that is being provided. From a purely practical level, ISO has a negative effect on image quality. How that negative effect gets there mechanically is interesting, but not relevant to the question.

In fact none of it is semantics. All of it is technically correct. All of it directly affects the image quality of capture data. And until a photographer does understand the significance of it (though not necessarily the technical details) their photography will be limited by that lack of understanding.

You're making it overly complicated. We don't need to know the physics behind why the knife cuts, to know that it does.

Thank you.

 

[runnah]

What are you going on about? Yes of course it changes the exposure. Because I am EXPOSING the film/sensor for a different amount of time based on the desired EXPOSURE.

That is what exposure means.

Yes, a "different amount of time". Changing the film didn't change the time. When you changed the shutter speed it change the time, and thus the exposure.

"Well I think that this chart blah blah shows that the blah blah."

Of course it makes a difference you knob! Do you honestly think that you discovered a big secret?

 

[apaflo]

In fact none of it is semantics. All of it is technically correct. All of it directly affects the image quality of capture data. And until a photographer does understand the significance of it (though not necessarily the technical details) their photography will be limited by that lack of understanding.

You're making it overly complicated. We don't need to know the physics behind why the knife cuts, to know that it does.

Ask any professional chef about knives. Ask any professional butcher about knives.. They won't agree with you about knives any more than I'm going to agree about what affects photography. You don't have to know about the physics, unless you want to be better at photography tomorrow than you are today!

Why do you feel it is necessary to denigrate what you don't care about or understand? Your lack of interest or understanding is not what does or not make something significant. I don't understand vast amounts of what Pablo Picasso thought was important about making art or images. I don't deride him or his knowledge.

 

[apaflo]

What are you going on about? Yes of course it changes the exposure. Because I am EXPOSING the film/sensor for a different amount of time based on the desired EXPOSURE.

That is what exposure means.

Yes, a "different amount of time". Changing the film didn't change the time. When you changed the shutter speed it change the time, and thus the exposure.

"Well I think that this chart blah blah shows that the blah blah."

Of course it makes a difference you knob! Do you honestly think that you discovered a big secret?

This is, afterall, a discussion forum. If you are not interested in the discussion why not just hike along until you find a discussion you are interested in. There is nothing to be gained from posting flame bait...