comparing lens + sensor size combos for low light

[Jerry R]

basically agree although I might phrase some things a little differently.

to answer your questions, yes to the first and no to the second.

 

[lumigraphics]

The equivalent aperture tells you what aperture on a full frame lens would give the same depth-of-field and the same total light as the one you're assessing.

The whole "total light" business is WRONG WRONG WRONG.

Remember that image size changes, so that your "4 times" is spread across a larger area to create the same image on the sensor.

There are other reasons that a larger sensor gives better IQ, but the total amount of light hitting it is NOT the reason.

 

[Abe R Ration]

Abe the bottom one is obviously has a lower S/N.

We agree

Was the bottom image normalized with pixel sharing?

Not sure what you mean by "pixel sharing".

The bottom image was resampled to to match the same output size with the upper image. (It's good to remember that practically all output images are resampled some way to the desired output size.)

I posted the images to demonstrate that increasing sensor area improves SNR of the output image even if the pixels are identical.



There is a very clear simulated version on the bottom of this page .

Abe R Ration, post: 55932426, member: 349459"]

Jerry R, post: 55932426, member: 349459"]
I posted the comparison from DXomark based on pixels not print because print DXomark down sizes which combines pixels which increase the S/N, effectively larger pixels.

Please view these two images at full size:

The bottom image was taken with crop factor of approximately 4,3. Identical pixels and exposures. The bottom image was normalized to the same output size with the top image. The top image is at "100%" or "1:1" scale, i.e. no combining of pixels were made.

Would you say that the SNR is the same or different? And why?

More on the topic here .

--
Abe R. Ration - amateur photographer, amateur armchair scientist, amaterur camera buff
http://aberration43mm.wordpress.com/

--
Sony R1, NEX C3 & 5R + Zeiss 24mm, 16-70, & FE 70-200 Lenses, Nikon V1 + 10-30 & 30-110 lenses.

--
Abe R. Ration - amateur photographer, amateur armchair scientist, amaterur camera buff

Abe R. Ration's lens and camera blog

This blog is about camera and lens technology and related sciences

aberration43mm.wordpress.com

[/QUOTE]
[/QUOTE]

 

[Jerry R]

I think I understand the images.

The top image contains more of the same pixels as the bottom.

The bottom is cropped from the full sensor and the top is taken with a FOV the just covers the image.

I think we are saying the same thing differently. I would say there are more pixels in each of the color squares of the top picture vs the bottom picture resulting in a higher S/N ratio.

Save the two pics it would help limit this type of discussion in the future.

 

[FingerPainter]

The equivalent aperture tells you what aperture on a full frame lens would give the same depth-of-field and the same total light as the one you're assessing.

The whole "total light" business is WRONG WRONG WRONG.

Remember that image size changes, so that your "4 times" is spread across a larger area to create the same image on the sensor.

There are other reasons that a larger sensor gives better IQ, but the total amount of light hitting it is NOT the reason.

 

[tomtom50]

The easy way. For modern sensors fill factor is high and read noise low so here's as easy rule of thumb:

1" & m43 & APS-C & FF are all about one stop apart. So in the OP example 2.8 on m43 is about the same as f4 on APS-C for low light.

To be more exact look up your sensor's ISO rating on Dxomark. A doubling of ISO is one stop. Even though modern sensors are close some are especially good for low light, and also each sensor size is not precisely twice as large or small.

 

[Abe R Ration]

So I will start the question with an example. Given these two combinations with premium type lenses:

a6000 type camera (APS-C sensor) + 1670z @ F4

vs

Om-D EM EM5 type camera (m43 sensor) + 12-40 F2.8

The m43 sensor is half the size of the APS-C sensor, and typically the larger sensors are better for reduced noise in shadows and low light conditions for the same generation.

First, here you can find pretty much all the relevant information regarding this.

APS-C is about 1,64 times later in area than M4/3, not twice as large.

Sorry, yes you are correct. m43 is half the size of a full frame sensor, not half the size of APS-C.

M4/3 is about 1/3,8 of the size of full frame, (26%) not 1/2

The relevant crop factor between these formats is about 1,3. The easiest way to compare performances of these two formats is just to multiply the smaller formats aperture number by 1,3, thus f/2,8 on M4/3 approximaely equals f/3,6 on APS-C.

As the aspect ratios of the sensors are different one could get slighlty more accuracy by using image sensor areas instead of crop factor, but crop factor is very convinient.

However, the m43 lens is a much faster lens in respect to the size of the sensor.

If one does cross format comparison then lens speed depends not only about the f-number, but also the format size. There is les than a third of a stop difference in "speed" between the lenses mentioned on the formats mentioned.

Furthermore, the small sensor means that the DOF for the m43 sensor will be larger,

thus one can employ a smaller aperture and not worry about parts of the image being out of focus in situations where there is no specific focal point.

There is no "deep DoF advantage" for smaller formats - it's a myth.

Instead the more light you capture per time unit (by opening the aperture) the more shallow the DoF is. f/3,6 on APS-C and f/2,8 on M4/3 have approximately equal light collection ability (same "noise" is the other exposure parameters are the same) and DoF.

(Note the "noise" above is about formats - individual cameras can have differing performance curves from each other.)

Rather, I am more interested in how does one compare across formats. Presumably, the reason why the 1670z is F4 and the 12-40 is F2.8 is that designing a F2.8 for an APS-C sensor would result in a larger lens. Thus there is a trade off here.

The f/4 on APS-C is as fast as f/3,1 would be on M4/3.f/4 on one format does not equal f/4 on another .

The size of the format has surprisingly little influence on the size of the lens as long as the lenses for different formats have similar entrance pupil diameter (*) and angle of view properties.

OK, I didn't realise. My statement was based on observations of lenses of different camera systems - the m43 seems smaller, for example the 70-200 F2.8 olympus looks smaller than the 70-200 F2.8 Canon for full frame

Is there a 70-200/2,8 for Olympus?

Regardless, it's good to remember that the same focal length and aperture number create a different effect on different formats, thus normally one would like to compare 35-100mm f/2,8 of M4/3 to 70-200/5,6 full frame lens. A FF lens for the same role might be slightly longer, but not necessarily much at all bigger otherwise. Of course in practise "equal" lenses are often not available.

. I think omission of a mirror also pays a role.

Lack of mirror allows for smaller wide angle lenses regardless of format - good examples are the many Leica-M mount lenses.

For tele lenses a mirrorless lens might even be longer (tough not necessarily is).

But nevertheless, I am not a optics engineer and can't say I understand the ratio of lens size to sensor size.

For long lenses the aperture diameter and focal length are the biggest parameters - for example a 400/4 lens will have to have at least 10cm diameter front element - near the image sensor there is no such mandatory minimum and it doesn't make that much difference if the image sensor is M4/3 or FF.

When aperture is changed to keep the DOF t the same the noise seems to be the same across formats:

I've written some which includes that .

 

[DrLogic]

So I will start the question with an example. Given these two combinations with premium type lenses:

a6000 type camera (APS-C sensor) + 1670z @ F4

vs

Om-D EM EM5 type camera (m43 sensor) + 12-40 F2.8

The m43 sensor is half the size of the APS-C sensor, and typically the larger sensors are better for reduced noise in shadows and low light conditions for the same generation.

First, here you can find pretty much all the relevant information regarding this.

APS-C is about 1,64 times later in area than M4/3, not twice as large.

Sorry, yes you are correct. m43 is half the size of a full frame sensor, not half the size of APS-C.

M4/3 is about 1/3,8 of the size of full frame, (26%) not 1/2

The relevant crop factor between these formats is about 1,3. The easiest way to compare performances of these two formats is just to multiply the smaller formats aperture number by 1,3, thus f/2,8 on M4/3 approximaely equals f/3,6 on APS-C.

As the aspect ratios of the sensors are different one could get slighlty more accuracy by using image sensor areas instead of crop factor, but crop factor is very convinient.

However, the m43 lens is a much faster lens in respect to the size of the sensor.

If one does cross format comparison then lens speed depends not only about the f-number, but also the format size. There is les than a third of a stop difference in "speed" between the lenses mentioned on the formats mentioned.

Furthermore, the small sensor means that the DOF for the m43 sensor will be larger,

thus one can employ a smaller aperture and not worry about parts of the image being out of focus in situations where there is no specific focal point.

There is no "deep DoF advantage" for smaller formats - it's a myth.

Instead the more light you capture per time unit (by opening the aperture) the more shallow the DoF is. f/3,6 on APS-C and f/2,8 on M4/3 have approximately equal light collection ability (same "noise" is the other exposure parameters are the same) and DoF.

(Note the "noise" above is about formats - individual cameras can have differing performance curves from each other.)

Rather, I am more interested in how does one compare across formats. Presumably, the reason why the 1670z is F4 and the 12-40 is F2.8 is that designing a F2.8 for an APS-C sensor would result in a larger lens. Thus there is a trade off here.

The f/4 on APS-C is as fast as f/3,1 would be on M4/3.f/4 on one format does not equal f/4 on another .

The size of the format has surprisingly little influence on the size of the lens as long as the lenses for different formats have similar entrance pupil diameter (*) and angle of view properties.

OK, I didn't realise. My statement was based on observations of lenses of different camera systems - the m43 seems smaller, for example the 70-200 F2.8 olympus looks smaller than the 70-200 F2.8 Canon for full frame

Is there a 70-200/2,8 for Olympus?

No, its actually a 40-150mm 2.8, corresponding to 80 - 300 mm. If memory serves Panasonic has a 35 - 100 mm F2.8, which corresponds to a 70 - 200.

Regardless, it's good to remember that the same focal length and aperture number create a different effect on different formats, thus normally one would like to compare 35-100mm f/2,8 of M4/3 to 70-200/5,6 full frame lens. A FF lens for the same role might be slightly longer, but not necessarily much at all bigger otherwise. Of course in practise "equal" lenses are often not available.

Yes, the DOF will be different.

. I think omission of a mirror also pays a role.

Lack of mirror allows for smaller wide angle lenses regardless of format - good examples are the many Leica-M mount lenses.

For tele lenses a mirrorless lens might even be longer (tough not necessarily is).

But nevertheless, I am not a optics engineer and can't say I understand the ratio of lens size to sensor size.

For long lenses the aperture diameter and focal length are the biggest parameters - for example a 400/4 lens will have to have at least 10cm diameter front element - near the image sensor there is no such mandatory minimum and it doesn't make that much difference if the image sensor is M4/3 or FF.

When aperture is changed to keep the DOF t the same the noise seems to be the same across formats:

I've written some which includes that .

--
Abe R. Ration - amateur photographer, amateur armchair scientist, amaterur camera buff
http://aberration43mm.wordpress.com/

 

[Abe R Ration]

Is there a 70-200/2,8 for Olympus?

No, its actually a 40-150mm 2.8, corresponding to 80 - 300 mm. If memory serves Panasonic has a 35 - 100 mm F2.8, which corresponds to a 70 - 200.

Since you mentione the M4/3 lens aperture numbers and you're converting the focal lengths to FF equivalent, it's a good idea to convert the aperture numbers as well in this context. Thus 40-150/2,8 -> 80-300/5,6 and 35-100/2,8 -> 70-200/5,6.

Regardless, it's good to remember that the same focal length and aperture number create a different effect on different formats, thus normally one would like to compare 35-100mm f/2,8 of M4/3 to 70-200/5,6 full frame lens. A FF lens for the same role might be slightly longer, but not necessarily much at all bigger otherwise. Of course in practise "equal" lenses are often not available.

Yes, the DOF will be different.

And light collecting ability - these two properties are tied togethed. Change one and the other will change two. Interestingly in this context the size of the sensor is practically irrelevant (outsize of macro regime).

More on comparing camera formats here .

 

[DrLogic]

Is there a 70-200/2,8 for Olympus?

No, its actually a 40-150mm 2.8, corresponding to 80 - 300 mm. If memory serves Panasonic has a 35 - 100 mm F2.8, which corresponds to a 70 - 200.

Since you mentione the M4/3 lens aperture numbers and you're converting the focal lengths to FF equivalent, it's a good idea to convert the aperture numbers as well in this context. Thus 40-150/2,8 -> 80-300/5,6 and 35-100/2,8 -> 70-200/5,6.

Regardless, it's good to remember that the same focal length and aperture number create a different effect on different formats, thus normally one would like to compare 35-100mm f/2,8 of M4/3 to 70-200/5,6 full frame lens. A FF lens for the same role might be slightly longer, but not necessarily much at all bigger otherwise. Of course in practise "equal" lenses are often not available.

Yes, the DOF will be different.

And light collecting ability - these two properties are tied togethed. Change one and the other will change two. Interestingly in this context the size of the sensor is practically irrelevant (outsize of macro regime).

More on comparing camera formats here .

 

[lumigraphics]

The equivalent aperture tells you what aperture on a full frame lens would give the same depth-of-field and the same total light as the one you're assessing.

The whole "total light" business is WRONG WRONG WRONG.

Remember that image size changes, so that your "4 times" is spread across a larger area to create the same image on the sensor.

There are other reasons that a larger sensor gives better IQ, but the total amount of light hitting it is NOT the reason.

 

[Abe R Ration]

Is there a 70-200/2,8 for Olympus?

No, its actually a 40-150mm 2.8, corresponding to 80 - 300 mm. If memory serves Panasonic has a 35 - 100 mm F2.8, which corresponds to a 70 - 200.

Since you mentione the M4/3 lens aperture numbers and you're converting the focal lengths to FF equivalent, it's a good idea to convert the aperture numbers as well in this context. Thus 40-150/2,8 -> 80-300/5,6 and 35-100/2,8 -> 70-200/5,6.

Regardless, it's good to remember that the same focal length and aperture number create a different effect on different formats, thus normally one would like to compare 35-100mm f/2,8 of M4/3 to 70-200/5,6 full frame lens. A FF lens for the same role might be slightly longer, but not necessarily much at all bigger otherwise. Of course in practise "equal" lenses are often not available.

Yes, the DOF will be different.

And light collecting ability - these two properties are tied togethed. Change one and the other will change two. Interestingly in this context the size of the sensor is practically irrelevant (outsize of macro regime).

More on comparing camera formats here .

 

[Abe R Ration]

Maybe you should explain to me how "total light" means anything. Since you guys brought it up first.

Hi, maybe I can join in here as the discussion seems to be rather heated not productinve at all.



[IMG width="400px" alt="This shows the effect of "total light". The "noisier" output images is made with crop factor of 4,3. All the exposure settings were identical as were the pixels. The only difference is in the crop factor. "]https://aberration43mm.files.wordpress.com/2015/02/compared.png[/IMG]
This shows the effect of "total light". The "noisier" output images is made with crop factor of 4,3. All the exposure settings were identical as were the pixels. The only difference is in the crop factor.

Light by nature is noisy. It follows Poisson distribution, thus standard deviation (or noise) is the square root of the signal or number of photons collected. The more light you collect, the more noise there will be, but as the signal increases at faster rate, the signal to noise ratio (SNR) increases - and SNR is what we perceive as "noise".

More on noise, "total light" and format comparisons on many pages of this blog .

For example:

Some samples on the context of total light, including a simulated one here .

About f/2 producing different effect on different formats .

Comparing camera formats (thus "total light" is considered)



If you have any questions, I'm happy to answer.



--
Abe R. Ration - amateur photographer, amateur armchair scientist, amaterur camera buff

Abe R. Ration's lens and camera blog

This blog is about camera and lens technology and related sciences

aberration43mm.wordpress.com

 

[jackalopemaui]

The equivalent aperture tells you what aperture on a full frame lens would give the same depth-of-field and the same total light as the one you're assessing.

The whole "total light" business is WRONG WRONG WRONG.

Remember that image size changes, so that your "4 times" is spread across a larger area to create the same image on the sensor.

There are other reasons that a larger sensor gives better IQ, but the total amount of light hitting it is NOT the reason.

 

[semifast]

To get the equivalent aperture you multiply by the crop factor of 1.31, so the f/2.8 lens on M43 is equivalent to f/3.67 on APS-C, which means the M43 should be 1/4 stop better in low light and will have 1/4 stop less DOF.

So equivalence is not only applicable on DOF (which I knew), but also in terms of exposure? That is, if I have both cameras set at say F4 and at the same ISO, the picture will be darker in m43?

I suppose thats an easy question to answer if only I had a m43 + APS-C camera, or APS-C camera + FF one, I could take two pics and compare.

F2.8 is F2.8. The aperture equivalence pertains to DoF, and has nothing to do with exposure. Aperture is one third of the exposure triangle, Aperture, shutter speed and ISO.

For example, high end smart phones have itty bitty sensors but have "wide open" apertures such as F2.0. However, if you do a macro shot or portrait with them it will be impossible to get subject isolation as one would with a larger sensor. Heck.... tiny cell phone sensors make for some really good landscape shots if the light is good enough to avoid high ISO. That's because the DoF is massive by comparison, so it is easy to get everything in focus.

No, Pixel Pooper is correct. Read the DPReview article on equivalence. It's basic physics.

I believe this is the section you are referring to. Quote (http://www.dpreview.com/articles/2666934640/what-is-equivalence-and-why-should-i-care/2):

The thing that complicates matters is ISO. ISO ensures that, if you expose a sensor to a given light intensity for a given amount of time, then you will get a certain brightness in your final (JPEG) image. Because it's based on intensity of light, it means that ISO depends on F-number, not equivalent aperture. This means that, a Four Thirds camera with a 50mm f/2 lens at ISO100 should produce a JPEG of the same brightness as a Full frame camera with a 100mm f/2 lens at ISO100 and, set to the same F-number and shutter speed, even though its smaller sensor means it is receiving 1/4 as much total light.

ISO is useful, in that it means that the same set of exposures work across all cameras (and frankly, it'd get confusing, otherwise). However, it ends up disguising how much total light each system gets. Since the light intensity is the same (per square mm), the Full Frame camera will receive four times as much light as the Four Thirds camera, during those exposures, because it has four times the sensor area, all experiencing that same intensity.

And this means that, for the same shutter speed, F-number and ISO, the camera with the largest sensor will have more total light to measure. And, unless the large sensor is significantly worse than the smaller one, it will produce a cleaner, less noisy image. It's likely that the large sensor camera will be bigger, heavier and more expensive, but it should provide cleaner images.

The flip-side of this is that, if you can fit a faster lens to a smaller format sensor or use a slower shutter speed then you can match the total light available to the larger system and gain similar image quality. However, this only really works in low light, where you're limited by the availability of light. In bright lighting conditions, where you're more worried about highlights clipping than you are about noise swamping the shadows, you can't simply open the aperture up to match a larger sensor's total light - you'll just end up over-exposing.

Yes, exactly. And this statement:

The equivalent aperture tells you what aperture on a full frame lens would give the same depth-of-field and the same total light as the one you're assessing.

That doesn't change anything about exposure. F/2.8 is f/2.8 as far as exposure, no matter what the sensor size is.

The dof is different. The amount of noise is possibly different. The exposure is the same.

Having more "total light" doesn't change the exposure.

 

[semifast]

Yes. However, what is exposure? It's just how much light per unit area is used to draw the image. Different formats have different amount of area, thus the same exposure settings on different formats create a different image.

Thus naturally also to create the same image on different formata you have to use different exposure settings.

In short: exposure is a tool, a mean to an end not the other way around.

ISO is not relevant in this context - it's not an exposure parameter .

Abe R. Ration - amateur photographer, amateur armchair scientist, amaterur camera buff
http://aberration43mm.wordpress.com/

What ? ISO is not an exposure parameter ?

There's two things that determine exposure, amount of light, controlled by aperture and how long the shutter is opened, and the sensitivity of the "film", expressed in a digital camera as ISO setting.

It's definitely an "exposure parameter."

And, different formats don't use different exposure settings for the same image.

 

[Abe R Ration]

Yes. However, what is exposure? It's just how much light per unit area is used to draw the image. Different formats have different amount of area, thus the same exposure settings on different formats create a different image.

Thus naturally also to create the same image on different formata you have to use different exposure settings.

In short: exposure is a tool, a mean to an end not the other way around.

ISO is not relevant in this context - it's not an exposure parameter .

Abe R. Ration - amateur photographer, amateur armchair scientist, amaterur camera buff
http://aberration43mm.wordpress.com/

What ? ISO is not an exposure parameter ?

That is the case .

There's two things that determine exposure

Three things.

, amount of light, controlled by aperture and how long the shutter is opened, and the sensitivity of the "film", expressed in a digital camera as ISO setting.

No. Exposure is defined by scene luminance, aperture number and exposure time.

Wikipedia has a nice article on the subjct .

In photography, exposure is the amount of light per unit area reaching image sensor.

The ISO setting does not change the amount of light reaching the image sensor in any way, form or shape.

Also ISO does not change the sensors sensitivity.

It's definitely an "exposure parameter."

Not according to International Organization for Standardization.

And, different formats don't use different exposure settings for the same image.

They do. I think I should clarify things:

First, when I say image I mean the image the lens draws into the image plane unless it is clear from the context I mean something else.

However this is often a bit of a shortcut as the size of the image is naturally different for each format. Often what is meant in this context is the information content of the image (i.e. we ignore the different sizes of the images - that is not relevant in this context).

So to be absolutely clear here:

The information content of the images are different on different formats if the same exposure settings are used, and vice versa, the information content can only be the same on different formats if different exposure parameters are used.

And the information content includes the depth of field and amount of light (and signal to noise ratio).

 

[Abe R Ration]

To get the equivalent aperture you multiply by the crop factor of 1.31, so the f/2.8 lens on M43 is equivalent to f/3.67 on APS-C, which means the M43 should be 1/4 stop better in low light and will have 1/4 stop less DOF.

So equivalence is not only applicable on DOF (which I knew), but also in terms of exposure? That is, if I have both cameras set at say F4 and at the same ISO, the picture will be darker in m43?

I suppose thats an easy question to answer if only I had a m43 + APS-C camera, or APS-C camera + FF one, I could take two pics and compare.

F2.8 is F2.8. The aperture equivalence pertains to DoF, and has nothing to do with exposure. Aperture is one third of the exposure triangle, Aperture, shutter speed and ISO.

For example, high end smart phones have itty bitty sensors but have "wide open" apertures such as F2.0. However, if you do a macro shot or portrait with them it will be impossible to get subject isolation as one would with a larger sensor. Heck.... tiny cell phone sensors make for some really good landscape shots if the light is good enough to avoid high ISO. That's because the DoF is massive by comparison, so it is easy to get everything in focus.

No, Pixel Pooper is correct. Read the DPReview article on equivalence. It's basic physics.

I believe this is the section you are referring to. Quote (http://www.dpreview.com/articles/2666934640/what-is-equivalence-and-why-should-i-care/2):

The thing that complicates matters is ISO. ISO ensures that, if you expose a sensor to a given light intensity for a given amount of time, then you will get a certain brightness in your final (JPEG) image. Because it's based on intensity of light, it means that ISO depends on F-number, not equivalent aperture. This means that, a Four Thirds camera with a 50mm f/2 lens at ISO100 should produce a JPEG of the same brightness as a Full frame camera with a 100mm f/2 lens at ISO100 and, set to the same F-number and shutter speed, even though its smaller sensor means it is receiving 1/4 as much total light.

ISO is useful, in that it means that the same set of exposures work across all cameras (and frankly, it'd get confusing, otherwise). However, it ends up disguising how much total light each system gets. Since the light intensity is the same (per square mm), the Full Frame camera will receive four times as much light as the Four Thirds camera, during those exposures, because it has four times the sensor area, all experiencing that same intensity.

And this means that, for the same shutter speed, F-number and ISO, the camera with the largest sensor will have more total light to measure. And, unless the large sensor is significantly worse than the smaller one, it will produce a cleaner, less noisy image. It's likely that the large sensor camera will be bigger, heavier and more expensive, but it should provide cleaner images.

The flip-side of this is that, if you can fit a faster lens to a smaller format sensor or use a slower shutter speed then you can match the total light available to the larger system and gain similar image quality. However, this only really works in low light, where you're limited by the availability of light. In bright lighting conditions, where you're more worried about highlights clipping than you are about noise swamping the shadows, you can't simply open the aperture up to match a larger sensor's total light - you'll just end up over-exposing.

Yes, exactly. And this statement:

The equivalent aperture tells you what aperture on a full frame lens would give the same depth-of-field and the same total light as the one you're assessing.

That doesn't change anything about exposure. F/2.8 is f/2.8 as far as exposure, no matter what the sensor size is.

If I may step in here - yes the above is right of course, but what is exposure?

To quote wikipedia:

In photography, exposure is the amount of light per unit area reaching image sensor.

Exposure is a useful tool, that's it. It by itself does not tell anything about noise levels or depth of field or other such metrics. The same exposure always creates a different effect for the output image on different formats. Always.

The dof is different. The amount of noise is possibly different. The exposure is the same.

DoF and amount of noise are always differen if the exposure is the same and format is different. Noise (or standard deviation) of light is the square root of signal (or number of photons). The more light you collect, the higher the signal to noise ratio (SNR), or what we perceive as "noise".

As exposure is a "per area" metric, having more area means that more light is collected by the same exposure and if more light is collected the "noise" is lower.

Having more "total light" doesn't change the exposure.

True (assuming it's the format size which changes the total light), but why is that relevant?

For more infomation on comparing formats you can go here .

 

[semifast]

Yes. However, what is exposure? It's just how much light per unit area is used to draw the image. Different formats have different amount of area, thus the same exposure settings on different formats create a different image.

Thus naturally also to create the same image on different formata you have to use different exposure settings.

In short: exposure is a tool, a mean to an end not the other way around.

ISO is not relevant in this context - it's not an exposure parameter .

Abe R. Ration - amateur photographer, amateur armchair scientist, amaterur camera buff
http://aberration43mm.wordpress.com/

What ? ISO is not an exposure parameter ?

That is the case .

There's two things that determine exposure

Three things.

, amount of light, controlled by aperture and how long the shutter is opened, and the sensitivity of the "film", expressed in a digital camera as ISO setting.

No. Exposure is defined by scene luminance, aperture number and exposure time.

Wikipedia has a nice article on the subjct .

In photography, exposure is the amount of light per unit area reaching image sensor.

The ISO setting does not change the amount of light reaching the image sensor in any way, form or shape.

Also ISO does not change the sensors sensitivity.

It's definitely an "exposure parameter."

Not according to International Organization for Standardization.

And, different formats don't use different exposure settings for the same image.

They do. I think I should clarify things:

First, when I say image I mean the image the lens draws into the image plane unless it is clear from the context I mean something else.

However this is often a bit of a shortcut as the size of the image is naturally different for each format. Often what is meant in this context is the information content of the image (i.e. we ignore the different sizes of the images - that is not relevant in this context).

So to be absolutely clear here:

The information content of the images are different on different formats if the same exposure settings are used, and vice versa, the information content can only be the same on different formats if different exposure parameters are used.

And the information content includes the depth of field and amount of light (and signal to noise ratio).

--
Abe R. Ration - amateur photographer, amateur armchair scientist, amaterur camera buff
http://aberration43mm.wordpress.com/

Information content ? I'm talking about exposure, you know, if a picture is too dark, too light, or just right.

The same aperture, shutter speed, and film sensitivity, will give the same exposure, no matter what size the film is.

I don't care what any organization says, if you shoot with Kodachrome 25, and Tri-x pan 400, you are not going to get correct exposure if you don't adjust your camera for the proper film speed.

As far as I know, that's what the word "parameter" means.

 

[semifast]

To get the equivalent aperture you multiply by the crop factor of 1.31, so the f/2.8 lens on M43 is equivalent to f/3.67 on APS-C, which means the M43 should be 1/4 stop better in low light and will have 1/4 stop less DOF.

So equivalence is not only applicable on DOF (which I knew), but also in terms of exposure? That is, if I have both cameras set at say F4 and at the same ISO, the picture will be darker in m43?

I suppose thats an easy question to answer if only I had a m43 + APS-C camera, or APS-C camera + FF one, I could take two pics and compare.

F2.8 is F2.8. The aperture equivalence pertains to DoF, and has nothing to do with exposure. Aperture is one third of the exposure triangle, Aperture, shutter speed and ISO.

For example, high end smart phones have itty bitty sensors but have "wide open" apertures such as F2.0. However, if you do a macro shot or portrait with them it will be impossible to get subject isolation as one would with a larger sensor. Heck.... tiny cell phone sensors make for some really good landscape shots if the light is good enough to avoid high ISO. That's because the DoF is massive by comparison, so it is easy to get everything in focus.

No, Pixel Pooper is correct. Read the DPReview article on equivalence. It's basic physics.

I believe this is the section you are referring to. Quote (http://www.dpreview.com/articles/2666934640/what-is-equivalence-and-why-should-i-care/2):

The thing that complicates matters is ISO. ISO ensures that, if you expose a sensor to a given light intensity for a given amount of time, then you will get a certain brightness in your final (JPEG) image. Because it's based on intensity of light, it means that ISO depends on F-number, not equivalent aperture. This means that, a Four Thirds camera with a 50mm f/2 lens at ISO100 should produce a JPEG of the same brightness as a Full frame camera with a 100mm f/2 lens at ISO100 and, set to the same F-number and shutter speed, even though its smaller sensor means it is receiving 1/4 as much total light.

ISO is useful, in that it means that the same set of exposures work across all cameras (and frankly, it'd get confusing, otherwise). However, it ends up disguising how much total light each system gets. Since the light intensity is the same (per square mm), the Full Frame camera will receive four times as much light as the Four Thirds camera, during those exposures, because it has four times the sensor area, all experiencing that same intensity.

And this means that, for the same shutter speed, F-number and ISO, the camera with the largest sensor will have more total light to measure. And, unless the large sensor is significantly worse than the smaller one, it will produce a cleaner, less noisy image. It's likely that the large sensor camera will be bigger, heavier and more expensive, but it should provide cleaner images.

The flip-side of this is that, if you can fit a faster lens to a smaller format sensor or use a slower shutter speed then you can match the total light available to the larger system and gain similar image quality. However, this only really works in low light, where you're limited by the availability of light. In bright lighting conditions, where you're more worried about highlights clipping than you are about noise swamping the shadows, you can't simply open the aperture up to match a larger sensor's total light - you'll just end up over-exposing.

Yes, exactly. And this statement:

The equivalent aperture tells you what aperture on a full frame lens would give the same depth-of-field and the same total light as the one you're assessing.

That doesn't change anything about exposure. F/2.8 is f/2.8 as far as exposure, no matter what the sensor size is.

If I may step in here - yes the above is right of course, but what is exposure?

To quote wikipedia:

In photography, exposure is the amount of light per unit area reaching image sensor.

Exposure is a useful tool, that's it. It by itself does not tell anything about noise levels or depth of field or other such metrics. The same exposure always creates a different effect for the output image on different formats. Always.

The dof is different. The amount of noise is possibly different. The exposure is the same.

DoF and amount of noise are always differen if the exposure is the same and format is different. Noise (or standard deviation) of light is the square root of signal (or number of photons). The more light you collect, the higher the signal to noise ratio (SNR), or what we perceive as "noise".

As exposure is a "per area" metric, having more area means that more light is collected by the same exposure and if more light is collected the "noise" is lower.

Having more "total light" doesn't change the exposure.

True (assuming it's the format size which changes the total light), but why is that relevant?

For more infomation on comparing formats you can go here .

--
Abe R. Ration - amateur photographer, amateur armchair scientist, amaterur camera buff
http://aberration43mm.wordpress.com/

Why is proper exposure relevant ? Gee, I don't know..

I already said DOF would be different. I qualified "noise" being different because it's much more complicated than just a mathmatical equation.

1. Noise doesn't matter, what matters is apparent noise. Noise you can see. So the subject matters. A lot.

2. light recpetors are imperfect mechanical devices. Therefore they can be better, or worse. Generally, over time, they get better. So a brand new smaller senosr may produce less noise than a much larger sensor from 10 years ago.