How to measure ISO

On 2015-10-31 18:21, Alfred Molon wrote:
In article ,
Whisky-dave says...
fortunately that insanity is no more, as digital is consistent shot to
shot for the life of the camera.

doesn't help if you have more than one camera though.

The difference in ISO is actually not a problem, because the camera
automatically sets the exposure and in manual mode you can determine the
right exposure by judging the histogram.

I was just curious how ISO could be measured. Got some feedback, but
isn't there a scientific definition, e.g. "100 ISO is the camera
sensitivity for which at an illumunance of ... lux the camera chooses
an exposure of ..." - something like this.

https://en.wikipedia.org/wiki/Film_s...ing_film_speed

https://en.wikipedia.org/wiki/Film_s...posure_i ndex

At the bottom of the article are various links to standards.

On 11/2/2015 6:49 AM, Whisky-dave wrote:
snip


by most yes.
It's like the old days countign the seconds while the enlarger was one exposing the paper to light. Countign to 30 seconds it never really mattered iof you were a second or two out even 5 seconds you'd hardly notice.



If you were doing color it mattered very much.
Some of us had clocks that precisely shut off the enlarger.



--
PeterN

On 2015-11-02 06:24, Whisky-dave wrote:
On Friday, 30 October 2015 19:59:10 UTC, Alan Browne wrote:
On 2015-10-30 12:38, Whisky-dave wrote:
On Friday, 30 October 2015 16:06:15 UTC, nospam wrote:
In article
,
Whiskers wrote:

Not a big problem, but the question is which camera
has the correct ISO setting, of if perhaps both
cameras deviate from the true ISO 100.

While you may have a point, it would not affect me to
know that the ISO setting in my camera is 100 off or
not. I know what I can expect from a setting, and set
the camera accordingly.

Things worked just as well in the days I set my exposure
to the Weston scale. The numbers don't really matter but
when you have more than one camera its nice if they share
a common scale.

exactly the point.

Using film, you could use the same batch of the same brand in
all your cameras - if you were worried enough about the
actual 'speed' you'd shoot a test strip for each batch of
film to determine the best 'speed' setting for that batch for
you. You'd also be aware of differences between the cameras
and lenses - and adjust shutter and aperture settings
accordingly, or the film speed setting if the camera had its
own 'automatic' exposure meter.

one of the downsides of film is that every batch is different,
one of numerous flaws that digital solves.

but batches of sensors are differnt too, which is whwere teh
original Q came from in that two cameras set set to ISO 100
should expose at exactly the same level of brightness, but they
don't.

If they're the same model camera I'd expect them to be too close
for casual measurement no matter what the batch. (Because I'd
expect the manufacturer to store calibration data on the sensor).

Expect what you want, does't make it happen.

Based on manufacturing experience with sensors of various kinds. As
batch to batch (or even unit to unit) sensor performance varies,
calibration values are often programmed onto the sensor part for reading
by the control or signal processor.



s
fortunately that insanity is no more, as digital is consistent
shot to shot for the life of the camera.

doesn't help if you have more than one camera though.

It's not so big a difference that you can't work them both.

yes I know that's why the term guide is used.

The ISO 12232:2006 standard



So hom come the OP was asking why his ISO one one camera is
difernt from teh other. You'd have thopugh with 10 bit minium
resolution they could get it right wouldn't you.

For a given model they should be too close for casual measure. And
in a precise test I'd assume far less than 1/3 stop.

yes in most cases when the camera manufacture is buying a particuar
sensor from a supplier but when two differnt suppliers are involved
making differnt sensor it isn't quite that easy.

Configuration control should mean the firmware knows which sensor
variant is installed and can adjust accordingly - at least for things
such as ISO.




ISO 'speed' ratings are guides rather than absolute
measurements. For electronic sensors there are at least three
different methods of deciding what the 'speed' of a sensor
is, and all allow for some subjective judgment. Wikipedia is
a good place to start reading.

the fact that there's more than one way to measure it makes
the rating rather useless.

So why have it ? Are you expecting the IOS to be measured at
5600k or 5400k or 3200k or anything else, or perhaps red, blue,
green, or yellow magenta or cyan light would be best. Moonlight
might be OK too for hi ISO.

The standard sets an objective. Engineers at each co. interpret it
as they see fit.

exactly as they see fit.

There is no exactly. Marketing may want engineering to fudge the
numbers so that "popular" tests (eg: dpreview) end up with more pleasing
results (eg: noise).


it's at best, a rough guide.

Yep. I remebr on boxes of film the sped was always given a s a
guide rathe rthan a dead cert, a bit like setting a gues make or
temerature on an oven. Yes I've even doen that and if I set my
oven to 180C I can get temeratures of 165-200.

Oven controllers are not very accurate - I agree with the range you
show there.

Running a lab which looks at what effect tolerancies have on
electrical ciruits can be important, our normal toleriences are 5%
that is Ok for almost all ciruits especailly digital but for precise
analogue mearsements we have 1% 0.1% and even 0.01%.

You won't get 1% analog control in a consumer oven. ±25°F is about as
good as one should expect. So 350°F gives 325 .. 350. That's a 3% error.

Alfred Molon wrote:
I have two cameras. At the same ISO, aperture and exposure time one
takes brighter images than the other one.

How to measure the real ISO values at which the cameras took the
pictures?



A similar situation existed years ago with mechanical shutters, which
oftentimes did not deliver their higher speeds with any degree of
accuracy. In digital, it is so very easy to make an occasional dial
adjustment, take more pictures, and get on with your life.

Mort Linder

In article , Mort wrote:

I have two cameras. At the same ISO, aperture and exposure time one
takes brighter images than the other one.

How to measure the real ISO values at which the cameras took the
pictures?


A similar situation existed years ago with mechanical shutters, which
oftentimes did not deliver their higher speeds with any degree of
accuracy. In digital, it is so very easy to make an occasional dial
adjustment, take more pictures, and get on with your life.

maybe crappy mechanical shutters did, but quality shutters were within
1/6th stop or better across all speeds, usually 1/10th stop.

In article , Alan Browne
wrote:

The standard sets an objective. Engineers at each co. interpret it
as they see fit.

exactly as they see fit.

There is no exactly. Marketing may want engineering to fudge the
numbers so that "popular" tests (eg: dpreview) end up with more pleasing
results (eg: noise).

in other words, exactly as they see fit. it's whatever they want it to
be that shows the product however they want it to be shown.

the real fun is when they detect when it's being tested and produce
results that don't reflect reality. it's not just volkswagen that plays
those games.

In article ,
Whisky-dave wrote:


Yes in other words you as a manufacter can chosee any 1 of 5 stanbdards to
follow.

which means there's no standard.

In article ,
Whisky-dave wrote:

because of that, many photographers bought film in large quantities so
that they could calibrate their workflow for one batch.

Some of that was to save money and always have film in stock.
I used to buy 50ft reels of HP3 , 100ft reels of HP4.
I still have the bulk film loader somewhere.

I'm not to sure that had such a thing as workflows then.

of course they did.

In what way, you'd ask the developer person to make up 10 pint of developer
he'd get the chenicals on kitchen scales and mill with quarts of water.

the amerture would most likely need to be more accurate due to scale.

whoosh.

they had to
freeze most of it so that it didn't expire before they had a chance to
use it. crazy!

No most didn't freeze it they kept it cool in the fridge.

that depends how long they're going to keep it.

freezing could damage the emulsion because of humidity, only the
biggest companies would freeze film, or thos ethat wanted to bulk buy
their film for 6 months or more.

no.

nospam wrote:
maybe crappy mechanical shutters did, but quality shutters were within
1/6th stop or better across all speeds, usually 1/10th stop.


Fiction, sir. I used Rolleiflexes, Nikon S-3 and SP, and mechanical
early SLRs starting in the 1950s, and their shutters were nowhere near
as accurate as you claim; nor were they "crappy mechanical shutters",
but rather Compur, Nikon, and other fine products.

Mort Linder

In article , Mort says...

A similar situation existed years ago with mechanical shutters, which
oftentimes did not deliver their higher speeds with any degree of
accuracy.

Ehhpp... exposure times are inaccurate? Even nowadays with digital
cameras? Is there any place where I can read more about this?
--
Alfred Molon

Olympus E-series DSLRs and micro 4/3 forum at
http://tech.groups.yahoo.com/group/MyOlympus/
http://myolympus.org/ photo sharing site