Bifurcator wrote:
You make it sound as if the photons have an awareness level and decide to spread themselves out when they see a larger sensor .
alundeb wrote:
The focal length does that.
Then how come the exposure values don't match your ideas here? If one exposes longer it's over-exposed... it doesn't get a better image quality as you seem to be suggesting.
This really is something new you seem to be saying here and doesn't match any of the evidence nor any of observations we've all made here over the past 2 or 3 years.
Given the light entering a 100mm f/1 lens = X the exposure values will be identical or nearly so no matter the sensor size.
Bifurcator wrote:
Then how come the exposure values don't match your ideas here? If one exposes longer it's over-exposed... it doesn't get a better image quality as you seem to be suggesting.
This really is something new you seem to be saying here and doesn't match any of the evidence nor any of observations we've all made here over the past 2 or 3 years.
Given the light entering a 100mm f/1 lens = X the exposure values will be identical or nearly so no matter the sensor size.
What am I missing?
I don't understand your question.
Baseline: u43 ISO 200, FF ISO 800.
If I want better image quality, I cannot get that with better exposure on u43 if the sensor only goes to ISO 200. But I can get it on the FF by lowering the ISO. Then of course I am out of the equivalent regime.
What I say has been said previously by other members as theSuede, Makten, Jman13 to mention some. theSuede also formulated a brilliant post about how the ligth gathering really is directly connected to the DoF.
I think it has something to do with focal length vs sensor size. You wont get same amount of photons if you use same 100mm f2.8 lens and full-frame vs m4/3s.
This is (at least I think) why compact cameras have such problems with not having enough ligth even when they have f2.8 or faster lens.
Simply put, I think that in photography its usually "bigger = better".
The best way to think of it is this way. If you have the same exposure, you're having the same amount of light per unit area. For a full frame sensor, you have four times the area, and thus four times the total amount of light. At a per unit area level, there's no difference, but when you make the final image, you are enlarging the m4/3 shot more (by a factor of sqrt 2 linearly, 4x in area), so you are using 4x less light to make the final image (if the sensor technologies are the same). Of course, whether this is perceptible in the final image will largely depend on size, ISO and keenness of eyes.
All I know is that if put a 100mm F2.8 lens (set to f/2.8) on a µ4/3 and point it at my white wall the exposure says 1/200 (for example).
If I take the same lens and put it on a FF camera and point it at the same wall (given the light hasn't changed) the exposure says 1/200 - exactly the same.
Is anyone saying that's wrong? And if they are could you show it? I just tried it with a 3 film cameras and my µ4/3 and it's indeed the same. Exactly the same. All were set to ISO 100.
OK I see, you're basically proving that sensor noise is directly related to sensor size.
Yep, that's fairly well known. But I don't think you can say it's the same total light (in your last sentence). This is what was throwing me. Obviously you're amplifying the less amount of light (from the smaller aperture) by setting the higher ISO sensitivity.
Bifurcator, you are confusing what we are saying by 'total light' with 'exposure' which are not the same. Exposure is light per unit area. TOTAL light, is the total amount of light gathered by the sensor.
Think of it this way...you've got a bucket and it's raining. The amount of rain is the same over all area for the same aperture, that is...if it rains 1", there will be 1" in your bucket regardless of the size of the bucket. This is exposure.
If, however, your bucket is 4 times larger...it's still 1" deep in the rain, but you have 4 times the amount of water in your bucket because the bucket is bigger.
You use all that water to create the image, so a full frame image at f/2 will use 4 times the amount of light to create the image as a m4/3 image at f/2. They will have the same exposure, but the S/N ratio is much higher on the FF chip because it uses more total light.
Again, given the same sensor technology, since as things have progressed, sensors have become more efficient at gathering light over the same area.
Bifurcator wrote:
OK I see, you're basically proving that sensor noise is directly related to sensor size.
Yep, that's fairly well known. But I don't think you can say it's the same total light (in your last sentence). This is what was throwing me. Obviously you're amplifying the less amount of light (from the smaller aperture) by setting the higher ISO sensitivity.
You get less light per square millimeter, but as the larger sensor has a greater surface you end up capturing the same # of photons.
Yeah. That sure seems like a convoluted way of saying things. A good cause of confusion. Just say there two ISO levels of IQ difference between a given sensor and one 1/2 its size. Why is it useful to talk about total number of photons in this way?
Bifurcator wrote:
Why is it useful to talk about total number of photons in this way?
As an engineer, that is a very natural thing to do. Sorry
It is just that we hear from time to time that 75 mm f/2 on u43 gathers the same light as 150 mm f/2 on FF with the same ISO and shutter speed. When we try to explain whay that is wrong, the world explodes for those who think that a smaller camera gives a shutter speed advantage. Makten was once banned on another forum for explaining it.
alundeb wrote:
As an engineer, that is a very natural thing to do. Sorry
It is just that we hear from time to time that 75 mm f/2 on u43 gathers the same light as 150 mm f/2 on FF with the same ISO and shutter speed. When we try to explain whay that is wrong, the world explodes for those who think that a smaller camera gives a shutter speed advantage. Makten was once banned on another forum for explaining it.
It's not wrong tho. You've created a confusion by talking about number photons over the surface of different sized sensors. A 75/2 on µ4/3 will indeed meter the same as a 150/2 on a FF. This is an unarguable fact. In fact if all lenses in a test group are perfect (without vignetting and etc) a 2mm f/2 lens will meter the same as 2000mm f/2 and on any camera you want to place it on. No one cares and no one can think about how many photons are required to make the image. We're photographers not photon counters.
Now as a function OF THE SENSOR SURFACE AREA (not yelling, just stressing the tense) one is using more information to form the image than the other. Thus what you have is sensor efficiency differential and nothing to do with the light gathering properties of a lens - which is what you actually hear people saying - and they're right. It's just that smaller sensors suck more. Hehe..
See, I knew there was something hokey going on here.
Bifurcator wrote:
No one cares and no one can think about how many photons are required to make the image. We're photographers not photon counters.
The noise in the image is a DIRECT function of the number of photons that went into the making the image (the photons that were converted to electrical charge in the sensor). For ISO below about 3200, NOTHING ELSE MATTERS FOR THE NOISE
True. No question that large sensor have a noise advantage! That's a given.
But small sensor users still get a perceived focal length to shutter speed advantage.
Consider that it could also be said that if the FF sensor had the same pixel density as the small sensor and a crop mode that matched the size of the smaller sensor then they too could get the same kind of FL/SS advantage. Or even just by cropping the image after capture. The FF sensor does not usually (D800?) have such a density tho. Se we small sensor users do get some advantage. Sure it's only FOV, crop factor or whatever but it's there. And IQ noise wise, is not really humanly detectable till us small guys hit ISO 400 or so.
alundeb wrote:
There must be at least 17 posts on this page only that explains how exposure and gathered light per image is not the same.
So, if I'm following this correctly, this is about the loss of efficiency with digital sensors at brighter F-stops and it means also that smaller sensors have a greater loss than larger sensors at the same F-stop (because, if the MP are the same then the sensel size must be smaller and less efficient with the smaller sensor)? If so, this is maybe what Fuji was stating with regard to choosing F2 for the speed of the X100 lens here:
"Designing an F1.6 or F1.8 lens is not so difficult; however, in the case of a digital camera, even if an aperture larger than F2 is used, the light receiving elements on the sensor cannot effectively use the brighter portion of the incoming light because of low incident light gathering efficiency."
Click "Story" then "Fujinon lens" at the top to see the quote.
Bifurcator wrote:
True. No question that large sensor have a noise advantage! That's a given.
But small sensor users still get a perceived focal length to shutter speed advantage.
Consider that it could also be said that if the FF sensor had the same pixel density as the small sensor and a crop mode that matched the size of the smaller sensor then they too could get the same kind of FL/SS advantage.
I'm in again
The perceived advantage is BS because we can just raise the ISO on the larger sensor camera and get the same shutter speed with a slower lens, and the same noise in the image. Back to start.
alundeb wrote:
"The term that needs a little clarification is "light gathering". We should be a little more specific on when we mean "light gathering per sensor area" or "light gathering per image".
Light gathering per sensor area is exposure. Light gathering per image is exposure*sensor area."
I'm not sure that is correct.
I seem to recall that "exposure" is the same as amount of light used for the whole image. Light gathering per sensor area is --> "exposure values".
If I'm wrong, or if this already has been corrected somewhere above (I got lost with all the quotations) I apologize and don't mind getting corrected.
ROTFLMAO 
Or even just by cropping the image after capture. The FF sensor does not usually (D800?) have such a density tho. Se we small sensor users do get some advantage. Sure it's only FOV, crop factor or whatever but it's there. And IQ noise wise, is not really humanly detectable till us small guys hit ISO 400 or so.