06-16-2016, 08:30 PM
(This post was last modified: 06-16-2016, 08:32 PM by wim.)
Quote:This explanation targeted the ISO discussion (ISO 100 (MFT) is not identical to ISO 100 (FF)).
It wasn't directed to you if you understood that already. :-)
A problem with the linear discussion stream.
This is not correct. Iso 100 is iso 100. If that weren't the case, you'd need a different light meter for every different sensor size and/or film format. That the one sensor has 4x the noise compared to the other at the same iso inder the same lightign conditions is a different matter altogether.
All you can state in this regard is that a 4/3 sensor is not the same as a FF sensor, noise wise, when images from these sensors are displayed or printed at the same size, with images taken of the same subject under the same lighting conditions - note that noise levels change with different lighting conditions as well, provided the sensor is capbale of capturing the different light levels, i.e., no under- or over-exposure. In order to get equivalent noise levels on a 4/3 as on a FF sensor, you need an exposure capturing 4 times the amount of light per square unit, or the same total amount of light. It just happens that iso is an easy way to manipulate this. At 100 iso you capture 4x the light than you do at iso 400. This is where this whole misunderstanding of "iso equivalence" originated from.
Kind regards, Wim
Gear: Canon EOS R with 3 primes and 2 zooms, 4 EF-R adapters, Canon EOS 5 (analog), 9 Canon EF primes, a lone Canon EF zoom, 2 extenders, 2 converters, tubes; Olympus OM-D 1 Mk II & Pen F with 12 primes, 6 zooms, and 3 Metabones EF-MFT adapters ....
Quote:Equivalent focal lengths are not the same focal lengths. They merely are focal lengths which give the same FOV.
Equivalent f-values are the same aperture sizes. They give the same DOF.
Equivalent ISO-settings are not the same ISO number, they merely give the same total light to form the image and result in similar exposure times.
How are ISO settings in digital decided upon by a manufacturer for a camera model? With a known light intensity, they take the whole sensor output in JPEG form with a chosen tonal curve and are free to choose a point on the resulting tonality curve which they deem appropriate, and call that for instance "ISO 400". They actually look at the whole sensor output.
Image noise does not count in this discussion, although it is a happy side effect. Why do I say that? We are not trying to get the same noise per camera, we are trying to get the same DOF and FOV in the first place. Only if somehow exposure time is somehow important (for instance, to freeze motion, or avoid camera shake), similar exposure times may also come into play. That is where equivalent ISO settings come into play.
Do we ever look for same noise results? No, we actually do not. It would be a silly mess, having to set things differently on a Nikon D300 than on a Nikon D70 or a Nikon D500, and never ending up with the same exposure time.
The happy coincidence of similar noise only occurs with sensors of more or less the same generation, using more or less similar technology.
Recap: Equivalent ISO settings are primarily about just one thing: trying to get similar exposure times. ISO settings are determined by the whole sensor output, not a square mm of it. Equivalent ISO settings result in the same amount of light captured by the sensor. Only with same generation, same technology sensors, equivalent ISO settings result in similar noise. If the sensors happen to have the same MP count, equivalent ISO settings result in similar amount of light captured per pixel.
Of course not. Similar exposure times are achieved by exposing the same way, i.e., same iso, same f-stop, same shutter speed or exposure time. there is no such thing in this equation as an "equivalent iso setting", only the same iso setting - provided the manufacturer follows teh standardized rules for is. Note that even with film those could and were manipulated to some degree. Not every 100 iso film was created the same, not every sensor base iso 100 is the same either.
I think you may not realize that there is a significant difference between total amount of light captured, and what you call iso equivalence. Obviously, a sensor that is 4x smaller than another, will capture 4x less total light than the other one, but iso, and related aperture and exposure times are still the same.
As to noise levels: funnily enough you do not need the same pixel count per sensor, all you need is a similar technology sensor to reach the same noise level, at the same amount of total photons captured, and even then only when an image is displayed or printed at the same size. This is why a 4/3 sensor is approximately 4 times noisier than a FF sensor, it just captures 4 times less total light for that same image surface area. And this is where the confusion with "iso equivalence" comes in, as the total amount of light can be made equal, namely by selecting a 4x slower iso for a 4/3 sensor (as compared to a FF sensor). IOW, noise is a function of total amount of light captured for a given surface area and magnification, while iso is a measurement for sensitivity, whether artificial or not.
To be very honest, IMO I think it is a moot point. Quality of sensors is at an extremly high level, much more so than film ever was, even with a 4/3 sensor.
Kind regards, Wim
Gear: Canon EOS R with 3 primes and 2 zooms, 4 EF-R adapters, Canon EOS 5 (analog), 9 Canon EF primes, a lone Canon EF zoom, 2 extenders, 2 converters, tubes; Olympus OM-D 1 Mk II & Pen F with 12 primes, 6 zooms, and 3 Metabones EF-MFT adapters ....
Quote:This is not correct. Iso 100 is iso 100. If that weren't the case, you'd need a different light meter for every different sensor size and/or film format. That the one sensor has 4x the noise compared to the other at the same iso inder the same lightign conditions is a different matter altogether.
All you can state in this regard is that a 4/3 sensor is not the same as a FF sensor, noise wise, when images from these sensors are displayed or printed at the same size, with images taken of the same subject under the same lighting conditions - note that noise levels change with different lighting conditions as well, provided the sensor is capbale of capturing the different light levels, i.e., no under- or over-exposure. In order to get equivalent noise levels on a 4/3 as on a FF sensor, you need an exposure capturing 4 times the amount of light per square unit, or the same total amount of light. It just happens that iso is an easy way to manipulate this. At 100 iso you capture 4x the light than you do at iso 400. This is where this whole misunderstanding of "iso equivalence" originated from.
Kind regards, Wim
This is what I wanted to express - ISO 100 (MFT) is not identical to ISO 100 (FF) because the noise level (aka amplification) is different.
Quote:Of course not. Similar exposure times are achieved by exposing the same way, i.e., same iso, same f-stop, same shutter speed or exposure time. there is no such thing in this equation as an "equivalent iso setting", only the same iso setting - provided the manufacturer follows teh standardized rules for is. Note that even with film those could and were manipulated to some degree. Not every 100 iso film was created the same, not every sensor base iso 100 is the same either.
I think you may not realize that there is a significant difference between total amount of light captured, and what you call iso equivalence. Obviously, a sensor that is 4x smaller than another, will capture 4x less total light than the other one, but iso, and related aperture and exposure times are still the same.
As to noise levels: funnily enough you do not need the same pixel count per sensor, all you need is a similar technology sensor to reach the same noise level, at the same amount of total photons captured, and even then only when an image is displayed or printed at the same size. This is why a 4/3 sensor is approximately 4 times noisier than a FF sensor, it just captures 4 times less total light for that same image surface area. And this is where the confusion with "iso equivalence" comes in, as the total amount of light can be made equal, namely by selecting a 4x slower iso for a 4/3 sensor (as compared to a FF sensor). IOW, noise is a function of total amount of light captured for a given surface area and magnification, while iso is a measurement for sensitivity, whether artificial or not.
To be very honest, IMO I think it is a moot point. Quality of sensors is at an extremly high level, much more so than film ever was, even with a 4/3 sensor.
Kind regards, Wim
Of course not. Similar exposure times can be reached in a multitude of ways. Stating otherwise is just nonsense, to be frank. You do not seem to try and understand what is actually written.
When talking about equivalency:
Equivalent focal lengths are not the same focal lengths. They merely are focal length which result in the same FOV.
Equivalent f-stops are not the same f-stops. They merely represent the same apparent aperture diameter, resulting in the same DOF.
Equivalent ISO settings are not the same ISO settings, nor are they the same number of photons per square mile/kilometer/inch or mm. They merely are ISO settings resulting in the same exposure time. As a side product, if sensors are from the same generation, using similar technology, one will have similar image noise.
That you in your own mind have fabricated a different idea of what "equivalent ISO setting" should mean, is your own business. But don't try to patronise me or others with that nonsense, when we are not wrong.
Please calm down, folks ... we are only talking about photography here - not about the end of the world.
We can always agree to disagree.
|
