Go Back   Steve's Digicams Forums > Digital SLR and Interchangeable Lens Cameras > Canon EOS dSLR

Reply
 
Thread Tools Search this Thread
Old Nov 30, 2004, 3:50 AM   #1
Member
 
Join Date: Jul 2004
Posts: 69
Default

How is ISO calculated for digital cameras? Because the noise at ISO 100 in a sunny day and noise at ISO 100 in the evening is way too different.
I know ISO is not about noise but everybody compares cameras by it...
"Whooa, my camera is noiseless at ISO 400, yours is grainy..." But what about the light...
10D vs. 20D vs. D7... How do they compare if not side by side? And even so, different lens apertures... different shutter speed... I am sure some could even claim camera fault comparing 20D vs. 20D at same iso by that.
I have a tendency to use ISO 400-800 in all the situations, sunny or cloudy, and for me 800 in daylight means lower noise than 400 in the evening and even faster shutter speed.
Of course, some cameras claim ISO 50... and they look like bait boxes :-)
pauza is offline   Reply With Quote
Sponsored Links
Old Nov 30, 2004, 5:51 AM   #2
Senior Member
 
Join Date: Jun 2004
Posts: 157
Default

I'm not going to try to answer all your questions but here is ont thing you should keep in mind.

Noise is a random scattering of red, green and blue pixels across an image. The amount of these pixels depends on the sensitivity the CCD or CMOS sensor is set to - the higher the sensitivity the greater the amount of 'noise' present.

If you take a picture at ISO 400 in daylight conditions there will be a certainamount of noise present, but the fact that the subject is in daylight with many varieties of colour will hide some of the noise and make it less conspicuous. A 400 ISO shot taken in the evening or at night is likely to have large dark or blackareas that will show up the noise in the image much more effectively.

Another thing to consider (but I am a less sure of my facts here)is that a 400 ISO shot in daylight is likely to result in a shutter speed between 1/100 and 1/2000 of a second (entirely dependent on light levels and aperture/shutter settingsat the time the shot is taken, of course). An evening/night shot at 400 ISO may require the shutter to stay open anywhere from 1/60 to4 or even 10 seconds. The amount of noise should accumulate the longer the shutter is openfor the sensor torecord the scene. So a 4 second night-time exposure should have far more noise then a 1/1000th daytime exposure.

Basically it comes down to what your personal taste is. If the noise present in an image is of less importance to you then achieving faster shutter speeds then by all means shoot at ISO 400-800. The general rule of thumb however is to shoot at the lowest ISO you are capable of shooting at that wont cause blurring due to slow shutter speeds.
redundo is offline   Reply With Quote
Old Nov 30, 2004, 7:18 AM   #3
Member
 
Join Date: Jul 2004
Posts: 69
Default

My question was:
How is ISO calculated for digital cameras?
And by far:
10D vs. 20D vs. D7... How do they compare (ISO and noise) if not side by side?

Sorry for the fillup. Thank you.
pauza is offline   Reply With Quote
Old Nov 30, 2004, 5:57 PM   #4
Senior Member
 
Join Date: Dec 2002
Posts: 5,803
Default

I believe ISO is a standard definition. The general rule goes like this:

The Sunny 16 rule says that a camera will exposure a sunny day properly at:
ISO100 1/125 f/16

And as you increase the ISO the shutter speed is 1/ISO value. (So technically the above shutter speed should have been 1/100, but you end up taking the closes shutter speed available which is usually 1/125th.)

But there are industry standards which they are supposed to test against which let them lable a setting at a specific ISO setting. I don't know how they do it, but that is how its done. I have heard from some people who have tested their cameras with the sunny 16 rule that their camera is often off by a bit. I don't know if this is because the manufacturer lied when labeling something as ISO100 (it was "close enough" so people will deal) or if there is a drift between specific cameras of a given model.

As for comparing noise.... I'm not sure what you're asking. There is no "official" amount of noise that makes something ISO100 or any ISO value. Noise is noise. It is there to some degree or not. This is why two different cameras will have different amounts of noise at the same ISO values. Because while noise is realted to ISO, technically ISO is completely independent of ISO. If you could make a noise free ISO800 or a ISO100 which was so noisy it looked like a snow storm.... that would be independent of it being either of those ISO values.

Eric
eric s is offline   Reply With Quote
Old Nov 30, 2004, 6:59 PM   #5
Administrator
 
Join Date: Jun 2003
Location: Savannah, GA (USA)
Posts: 22,378
Default

pauza wrote:
Quote:
My question was:
How is ISO calculated for digital cameras?
And by far:
10D vs. 20D vs. D7... How do they compare (ISO and noise) if not side by side?

Sorry for the fillup. Thank you.
Do you mean D7 (as in Konica-Minolta Dynax 7) or D70 (as in Nikon D70).

Some reviewers have compared noise levels on these models. The general concensus is that the new EOS-20D has lower noise levels than the EOS-10D or Nikon D70. But, it may not be enough of a difference for it to have any bearing on your shots (depending on the lighting conditions, the ISO speed you use, etc.).

Noise comparisons are always suspect, because one camera may do better in one conditon (lighting levels, aperture, shutter speed, etc.), and another better in other conditions. Exposure is sometimes very critical to noise levels, as it's not uncommon to see more noise in images if any areas are underexposed even a little bit at higher ISO speeds.Settings for image processing parameters can also come into play (for example: more sharpening in camera tends to accentuate noise).

As far as the new Konica-Minolta Dynax 7, I have seen one comparison where it was virtually neck and neck with the EOS-20D through ISO 3200 in a controlled test (and even had a slightly better signal to noise ratio on the charts at everything over ISO 100). But, the difference in them was really negligible for that one test. In another test with different conditions, the EOS-20D may do better. Chances are, in real world conditions, there is not going to be any appreciable difference between them.


JimC is offline   Reply With Quote
Old Dec 1, 2004, 7:02 AM   #6
Member
 
Join Date: Jul 2004
Posts: 69
Default

As far as I know ISO stands for an industry standard. Well, I don't think they abide to it, at least for the smaller cameras, but... the big ones are next in line...
As far as I heard, tested by iso rules (probably " The Sunny 16 rule"), Canon overexposes and KM underexposes, in similar conditions, same all, even similar lenses. (20D vs. 7D)

eric s wrote:
Quote:
As for comparing noise.... I'm not sure what you're asking. There is no "official" amount of noise that makes something ISO100 or any ISO value.
Well, it was just talking about comparing cameras. I am a bit concerned about giving a camera credits just for noise at a specific ISO, that's all. And test them in different light condition, as said by all.

And another thing... while you can "program shift"... why not getting same results at same ISO for digital cameras? ...and I am not talking about DOF gain...
pauza is offline   Reply With Quote
Old Dec 1, 2004, 7:42 AM   #7
NHL
Senior Member
 
NHL's Avatar
 
Join Date: Jun 2002
Location: 39.18776, -77.311353333333
Posts: 11,547
Default

pauza

I believe everyone missed your point which is well taken :idea:

The ISO (in this case sensitivity) is indeed a standard... and if you measure with any light meter with the same ISO setting all cameras will have the same shutter/aperture combination

What you are trying to bring about is the noise levels are different with different EV values -> noise @ 1/200s is less than noise @ 2s. At longer shutter speed the signal coming from the light source is weaker and a camera has to average this lesser signal over a longer period of time with the camera's internal noises even thogh the ISO setting hasn't changed!

-> I agree with you all cameras have to be tested the same way (same EV and same shutter speed... etc) and this is not standardized with the folks who tested them ; However the ISO is indeed a standard... As long as the testers stayed consistent between cameras it's OK to gather comparison data from. You can't compare a picture at one site vs another one at different site :?

FYI check page 3: http://www.i3a.org/pdf/debunking_specsmanship.pdf
NHL is offline   Reply With Quote
Old Dec 1, 2004, 9:30 AM   #8
Senior Member
 
Join Date: Nov 2004
Posts: 378
Default

Your question is "How is ISO calculated for digital cameras?" -- I hope this link directly answers that question:

From http://velatron.com/dca/ISO/

The photographic ISO film speed rating has a long and sordid past which included several different names such ASA and DIN and conversions between them. ISO is an acronym for the International Standards Organization which establishes and monitors many standards in many industries, products, and services and the specific standard which applies to film speed is ISO 5800:1987. Today most countries and manufactures use this ISO film speed standard and so the confusion has been alleviated, that is until digital cameras became of age and started to use ISO ratings.

With a traditional camera the film speed or ISO is a measure of the film's sensitivity to light and higher ISO values like 800 and 1600 are more sensitive to light where ISO 100 or 200 are less sensitive. In a digital camera there is no film but there is an image sensor, usually a CCD, which has an adjustable gain amplifier and thus a sensitivity setting. Early digital cameras either had no ISO settings, a sensitivity setting or were equivalent to ISO 100 or ISO 50 only. Newer models are offering a compliment of ISO settings and as an example the Nikon Coolpix 995 has ISO100, ISO200, ISO400 and ISO800 settings.

Back to traditional films for a moment. A traditional film's increase in ISO represented an increase in the size of the silver bromide crystals used to cover the film. The larger these crystals are the faster the film, the finer the crystals the slower the film. The photographic result of larger crystals or faster film is that the image has less resolution and appears grainier. A digital camera has a similar trade off. As the image sensor's amplifier is set to higher gain this increases the dark current in the collector which in turn, based on the length of the exposure, adds noise to the resulting image.

...

The ISO standard has very detailed specifications and all are related to chemical film and not digital cameras. Most digital camera manufactures refer to their ISO setting as ISO equivalent and so we will treat it as equivalent. Table 1 below shows the various shutter speeds at various ISO equivalents and is based on standard exposure formulae. By moving horizontally in our table we encounter equivalent exposures. So a 1/60 of a second exposure in ISO 100 is equivalent to a 1/480 of a second in ISO 800. Similarly, a 7.5 second exposure @ ISO 800 would result, theoretically, in the same exposure for 60 seconds at ISO 100.

(the table referred to above is available at http://velatron.com/dca/ISO/)

So your answer is that manufacturers attempt to label digital ISO to be ther rough equivalent of the amount of light able to be received by the CCD (or CMOS) as with whatever ISO film speed. This would be related to grain and sharpness, then, by saying that if the CCD or CMOS are "better" (that is, more sensitive to light) then less gain would be required to capture the image at that sensitivity, and thus the picture would have less grain.

I think the only way to truly compare grain at different ISO speeds is to compare cameras and images side by side in different lighting conditions.


perdendosi is offline   Reply With Quote
Old Dec 1, 2004, 9:35 AM   #9
NHL
Senior Member
 
NHL's Avatar
 
Join Date: Jun 2002
Location: 39.18776, -77.311353333333
Posts: 11,547
Default

This is the standard for ISO speed determination in electronic still camera (ISO-12232):
http://www.opsci.com/AppNotes/ShortO...talCameras.pdf

Notice on page 4: "Since the noise performance of an image sensor may vary significantly with exposure time and operating temperature,..."
NHL is offline   Reply With Quote
Old Dec 1, 2004, 1:32 PM   #10
Senior Member
 
Join Date: Dec 2002
Posts: 5,803
Default

I need to read that quote of the standards doc. Looks very interesting.

I was just going to throw out that I didn't think that what NHL talked about was the point pauza was trying to get at... but I do still agree with him. I am very sure that different cameras will have different noise characteristics at different shutter speeds and the same ISO value. I'd like to think there isn't that much of a difference (but again, that is camera dependent) but I'm sure that in some cases there is.

I know that neatimage recommends (or at least used to) that you have a different noise profile for every ISO and many different shutter speeds.

Technically, the temprature that the picture is taken at will also effect things as heat effects the sensor and the noise in the image. It probably doesn't matter too much until you get to extremes, but it is there. So I'd think that the same snowy picture taken in the artic will have less noise than one taken near Boston Ma.

Eric
eric s is offline   Reply With Quote
 
Reply


Thread Tools Search this Thread
Search this Thread:

Advanced Search

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 Off
Pingbacks are Off
Refbacks are Off



All times are GMT -5. The time now is 11:53 PM.