ISO question

[Badcast]

I read on a post not to use the inbetween ISO settings like 250 500 is there a reason for this or is this not correct.

Randy

[jiro]

Any specific link based on that? I use ISO 250 and ISO 320 on some of my shots and I still find the exposure OK. That is interesting. The only issue I see is that it would be hard to do mental calculations based on aperture and shutter adjustments due to the fractional increments.

[MiniChris]

When I introduce photography to my students, I explain to them that each exposure component has to be seen as a unit of light. An f/stop is a unit, a shutter speed number is a unit, ISO's are units as are +/- EV's and pushes. Technically, each degree of chemical temperature is a unit as is each increment of processing time. When one considers all the "perfect" exposure combinations achievable by adding in or subtracting out for exposure variables, it generally takes a pretty significant off course steer to change the outcome of an exposure. A 1/2 stop is miniscule in that regard.

Unless you were shooting straight manual, I don't you'd notice any exposure shifts, and even then, not sure the 1/2 or 1/4 stops these variations would represent would have a significant enough impact on the final outcome. It's really more of a tweaking than a real shift in exposure. Think of it like adding a 10% screen or multiply layer adjustment. Just isn't enough to lose sleep over.

[Colin Southern]

I read on a post not to use the inbetween ISO settings like 250 500 is there a reason for this or is this not correct.

Randy

Hi Randy,

It's nothing to do with exposure, but has to do with noise instead. When you change ISO by full stop amounts (100, 200 etc) this is accomplished in camera be changing the gain of the signal before it hits the Analog to Digital converter - however - when one changes by an odd-stop amount it's done two ways; firstly by changing the gain as "per normal" but then the partial stop adjustment is done my digital manipulation of the converted digital data, which amplifies noise.

So in theory, you'll get less noise at, say, ISO 800 than you will at ISO 500. Tests have backed this up for some Canon cameras, and yet I've also seen results that swung the other way slightly. Personally, I don't shoot off ISOs, but in reality, the two things you can do to avoid noise problems in the first place is (a) ensure the shot is exposed correctly in the first place (that one is the absolute golden rule #1 by a HUGE margin) (even if you have to use a much higher ISO -- you'll get far less noise using, say, ISO 1600 than you would under-exposing a shot by 4 stops @ ISO 100), and (b) Avoid cropping a high-ISO shot excessively (it doesn't increase the noise, but it can make it more visible).

In reality (as opposed to the theory), it's unlikely to make much of a difference with modern cameras.

Hope this helps

[Colin Southern]

PS: Keith did a little test here to back up what's beed written ...

http://www.northlight-images.co.uk/a...ds3_noise.html

[Badcast]

Thanks Colin, I shot a track meet yesterday and battled a heavy dark overcast toward the end . I found myself bumping up the ISO alot to keep up my shutter speed.When I loaded the images to view I had less noise at 800 than 500 or it sure looked that way. Then I remembered reading somewhere about it. I should have got closer and used a faster lens but I didnt want to be intrusive. My first track meet and dont want to burn any bridges. Thanks for answering my question.

Randy

[MiniChris]

Well, I would take exception to ISO having nothing to do with exposure, otherwise, why have it...? Go from a 100 ISO to 200 and you've gained a unit of light (an exposure value) regardless of the noise factor. Of course, even though there is a definite corrolation between grain and noise, I would agree that bumping up the ISO does increase the chance of noise more incrementally and I think it is more noticeable digitally than in film.

[Colin Southern]

Well, I would take exception to ISO having nothing to do with exposure, otherwise, why have it...? Go from a 100 ISO to 200 and you've gained a unit of light (an exposure value) regardless of the noise factor.

I've had "debates" along these lines before ... it all comes down to the definition of exposure - I'm considering it purely from the exposure triange perspective; with aperture constant, it you double the ISO, you halve the time the shutter is open - and thus the exposure remains the same (and - ignoring the decrease in DR at the higher ISO - the average value of the tones in the capture remains the same - as does the source light). But I know what you're saying too, which is also correct - we're just talking about different things.

In real terms, using a high ISO trades increased noise & loss of DR for potentially less image degradation due to camera shake and/or motion blur or insufficient DoF. The irony is that many people avoid high ISO settings like the plague when in reality the noise isn't visible at normal print and/or display sizes (ie just about anything other than 100% view in Photoshop!), so long as the 2 cardinal rules of not under-exposing and not over-cropping are followed; again, in real terms, high ISO noise is far far far far less damaging to an image than camera shake / motion blur / insufficient DoF.

I think it is more noticeable digitally than in film.

I don't think so - viewing a digital file on a monitor at 100% resolution - assuming a modern high MP sensor - is equivalent to blowing up a negative to over 50 inches wide by 30 something inches high - and I'll bet there wasn't much 1600 or 3200 ASA film you could do that to and have less visible noise than the equivalent ISO from a modern camera

[MiniChris]

You say tomatoes, I say tamaters...we are talking the same language and coming to differing interpretations, yet in the middle, it is still a tasty treat.

[Sahil]

Just FYI in Hindi (Indian National Language) Thats what Tomatoes are called, Tamater.

[Colin Southern]

You say tomatoes, I say tamaters...we are talking the same language and coming to differing interpretations, yet in the middle, it is still a tasty treat.

Ah yes, but are they a fruit or a vegetable?

[MiniChris]

They are a fruit.

[Colin Southern]

They are a fruit.

Not according to your supreme court! http://en.wikipedia.org/wiki/Nix_v._Hedden

Any bets on beans? - fruit or vegetable?

[MiniChris]

Peas and beans are considered legumes, which are neither fruits or vegetables...leagally speaking, a tomato may be called a vegetable but botanically, it is a fruit.

Did you know that a tuna is as different from a halibut as a pig is from a cow?

[revi]

Peas and beans are considered legumes, which are neither fruits or vegetables...leagally speaking, a tomato may be called a vegetable but botanically, it is a fruit.

Did you know that a tuna is as different from a halibut as a pig is from a cow?

Peas and beans are actually seeds iirc...
and speaking about legalese, in France a carrrot is a fruit (you can make juice with it ) and the French legume is the English vegetable. It all depends on your frame of reference.

[Black Pearl]

On the wall behind the counter at work we have thirty plus images printed at 15x10 inches. They have been taken with eight different cameras and at pretty much every ISO from 100 to 3200....one of them was taken with my iPhone.

If I stand a few inches from each I can just about tell the differences between them and can spot the very high ISO ones - If I stand a couple of foot back, in effect a normal viewing distance, I can't tell the difference between any of them. I may also point out that not a single customer who has commented on them has asked in a negative way what ISO they were taken on or if a particular one was taken on a Canon or a Nikon or a Sony (ad-infinitum) so while I understand why these things are debated on forums it really doesn't matter for images that are produced as real, printed photographs.

The moral of this....rant....is print your pictures, view them at normal distances and forget about pixel peeping.

[Colin Southern]

On the wall behind the counter at work we have thirty plus images printed at 15x10 inches. They have been taken with eight different cameras and at pretty much every ISO from 100 to 3200....one of them was taken with my iPhone.

If I stand a few inches from each I can just about tell the differences between them and can spot the very high ISO ones - If I stand a couple of foot back, in effect a normal viewing distance, I can't tell the difference between any of them. I may also point out that not a single customer who has commented on them has asked in a negative way what ISO they were taken on or if a particular one was taken on a Canon or a Nikon or a Sony (ad-infinitum) so while I understand why these things are debated on forums it really doesn't matter for images that are produced as real, printed photographs.

The moral of this....rant....is print your pictures, view them at normal distances and forget about pixel peeping.

Bravo Robin ... if I had a list titled "This person gets it", I'd be adding your name. I wish there were more people who understood this MP -v- ISO -v- viewing distance "truth" ... we'd hear a whole lot less about noise - high ISO modes - ridiculous print densities - ridiculous sensor MP counts ... and quite possibly start appreciating the art a whole lot more. Sadly though, with regards to these things, our mission as educators is far from complete

[RonH]

Crikey ... I thought my browser had redirected me to a gardening website Phew ... got a bit worried that I was hitting one of those 'give us your money or we will give you a virus' sites
But reading on ... looks like I'm still 'home'

[Chesil]

On the wall behind the counter at work we have thirty plus images printed at 15x10 inches. They have been taken with eight different cameras and at pretty much every ISO from 100 to 3200....one of them was taken with my iPhone.

If I stand a few inches from each I can just about tell the differences between them and can spot the very high ISO ones - If I stand a couple of foot back, in effect a normal viewing distance, I can't tell the difference between any of them. I may also point out that not a single customer who has commented on them has asked in a negative way what ISO they were taken on or if a particular one was taken on a Canon or a Nikon or a Sony (ad-infinitum) so while I understand why these things are debated on forums it really doesn't matter for images that are produced as real, printed photographs.

The moral of this....rant....is print your pictures, view them at normal distances and forget about pixel peeping.

Mostly I agree. But check out prints from Ansel Adams against a similarly sized print from an iphone and I think you'd see the difference. There are quality differences sometimes.

My view is that many photographers (including me) are geeks and lose sight of the final image. But beauty or magnetism of the final image is what hangs photographs on walls and not whether they were taken on an iPhone or a large format camera.

[Black Pearl]

Mostly I agree. But check out prints from Ansel Adams against a similarly sized print from an iphone and I think you'd see the difference.

You would but would you spot the difference between 400iso and 500iso?

I love the cutting edge of technology and I love gadgets but I love photography more and I sometimes get the feeling that tech is getting in the way of art.