Equivalence in f-stops Yes, NO or But...

[george013]

Hi Bill!

Thank you for your contribution. Yes, I am a great believer in practical demonstrations (seeing is usually believing) and one of the reasons I offered the much-maligned Northrop videos was because, flawed as they may be, they make some kind of effort to show the effects he is trying to express in practical terms.

I think a lot of the debate about f-stops in equivalence comes from a fundamental difference of understanding of what an f-stop is in terms of delivering light from the lens and what equivalence is trying to say regarding light gathering by the sensor. Much of that seems to boil down to an argument between light intensity per-unit-area, vs. light gathered by the whole sensor's area.

In that case people will talk past one another and the debate will rage forever!

Can you give a more specific example?

George

[Dave Humphries]

Surely a far better DoF test than a tight portrait (real world though that may be), would be to use one of those auto-focus accuracy checking devices. That should allow us to read off the numerical scale/tape measure exactly how much is 'in focus' and more easily compare the two shots.

[george013]

Surely a far better DoF test than a tight portrait (real world though that may be), would be to use one of those auto-focus accuracy checking devices. That should allow us to read off the numerical scale/tape measure exactly how much is 'in focus' and more easily compare the two shots.

I think this is just what Trev wanted to avoid. He's looking for an explanation.

I've been thinking about the same subject and I like figures. So for a start.



A simple drawing. Shown an angle of view. Any sensor size on the left will have the same framing, aov and distance to the subject. This is wat William did: 2 persons taking a shot of each other, one with a ff and one with a dx. Focal length of the ff was corrected with the crop factor of the dx, or the other way around.

Since the f-number is focal length/diameter aperture the f-number doesn't change. In the drawing the ratio between sensor size and focal length doesn't change.
What does change is the aperture diameter. For the bigger sensor has more surface, more light must be needed. That part is in the definition of the f-number: a bigger diameter of the aperture.

George

[Tronhard]

What I was saying at the beginning was that equivalence looks at the performance of photographic elements relative to a 35mm format.

Given that the sensors of APS-H are reduced by a factor or 1.3, APS-C by a factor of 1.5 or 1.6 and Micro 4/3 is reduced relative to 35mm format by a factor of 2, what is the significance of the performance of a the sensor in receiving light from the lens when used on these formats.

As I said it has been pretty well accepted that when considering the Field of View (FoV), and when considering the IMPACT OF THE CROP FACTOR ON THE IMAGE, one multiples the focal length as shown on the lens by the crop factor.

Thus a lens identified as a 100mm f4, the apparent focal length would be APS-H= 130mm, APS-C = 150 or 160mm and Micro 4/3 = 200mm. The physical configuration of the lens does not change, nor does what it projects through the aperture, but what is gathered by the sensor is - its FoV is cropped by the smaller light-gathering area.

By the same token the diaphragm will NOT physically be changed by mounting it on various-sensor cameras - the lens ITSELF is not physically altered in any way.

So here is the question that I have tried to express. IS the Equivalent (or if you will apparent) F-stop ALSO changed by the crop factor? The logic being that the physical diaphragm on the lens will not be altered, nor is what it projects, but the crop effect on what is picked up on the sensor will alter the resultant image.

Thus in the example above the apparent f-value of the captured image would change by crop factor.
APS-H would be f4 x 1.3 = f5.2
APS-C would be f6 or f6.4
Mi4/3 would be f8

Further the intensity of light transmission PER UNIT AREA would be constant, but because it would be falling on smaller targets, it would have a lower TOTAL light transmission - which is what would impact the total image brightness.

So, this is how it was presented in one submission on another forum:
"The aperture is specified as an F-number N, defined by the well-known formula

N = f/D

where f is the focal length of the thin lens (a.k.a. the effective focal length of the original lens assembly), and D is the diameter of the diaphragm opening." - quoted from lecture notes, Prof. LeVoy.

Note the word Effective, because if we place the lens on a crop sensor, it's effective (if you will apparent) focal length (fe) increases because its field of view narrows by the crop factor of the sensor, yet D remains a constant. So that changes the equation

To balance the equiation one have to increase both sides by that amount.

f x crop factor/D = Ne = N x crop factor

So for a Nikon APS-C sensor........... Ne = f4 x 1.5 = 6

for M4/3 ......................................... Ne = f4 x 2.0 = 8

where Ne and Fe are Equivalent values, not the original physical ones. The lens, its mechanical focal length and physical aperture have not changed, so the light intensity per unit area remains constant. In other words everything about the lens is still the same

What HAS changed is the ability of the sensor to gather total light to produce the image and thus image brightness.

This has been disputed by experienced photographers who say that they cannot accept that logic, hence the interesting dilemma! Is equivalence in f values real or not, and why?

Then there is the issue of the effect on DoF, but lets deal with the Fe issue FIRST please.

This is as specific example of the issue I am trying to get quality material for

[pnodrog]

Further the intensity of light transmission PER UNIT AREA would be constant, but because it would be falling on smaller targets, it would have a lower TOTAL light transmission - which is what would impact the total image brightness.

No. The apparent brightness is a function of unit area. The total energy is a function of overall area. Forget having a lens and think about an array of solar cells. The voltage generated by each cell is a function of brightness and the total power collected a function of area (number of cells).

[Tronhard]

Before you say anything more just re-read what I just wrote, because I had to do some serious editing, the whole thing turned to custard when I was posting it and I got quotes mixed in with comments mixed in with stuff from elsewhere.

So to be clear... I'm not sure what you are saying No to Paul... From what you are saying afterwards you SEEM to be saying that the premise outlined above i.e. that an equivalent F value is impacted by the crop factor because of the total light gathered by the sensor. Which is the suggestion put forward (despite my posting disaster!)

Either way I am seeking documentation because both sides are claiming logic (and no I'm NOT judging your opinion personally!)

[Tronhard]

WHAT IS THE MATTER WITH THIS SITE TODAY??? I get delays of up to 30 seconds for a post or edit and sometimes a time-out... is anyone else experiencing this????

[george013]

What I was saying at the beginning was that equivalence looks at the performance of photographic elements relative to a 35mm format.

Given that the sensors of APS-H are reduced by a factor or 1.3, APS-C by a factor of 1.5 or 1.6 and Micro 4/3 is reduced relative to 35mm format by a factor of 2, what is the significance of the performance of a lens when used on these formats.

As I said it has been pretty well accepted that when considering the Field of View (FoV), and when considering the IMPACT OF THE CROP FACTOR ON THE IMAGE, one multiples the focal length as shown on the lens by the crop factor.

Thus a lens identified as a 100mm f4, the apparent focal length would be APS-C= 130mm, APS-C = 150 or 160mm and Micro 4/3 = 200mm. The physical configuration of the lens does not change, nor does what it projects through the aperture, but what is gathered by the sensor is - its FoV is cropped by the smaller light-gathering area.

By the same token the diaphragm will NOT physically be changed by mounting it on various-sensor cameras - the lens IS not physically altered in any way.

So here is the question that I have tried to express. IS the Equivalent (or if you will apparent) F-stop ALSO changed by the crop factor? The logic being that the physical diaphragm on the lens will not be altered, but its effect on what is cast on the sensor will.

Thus in the example above the apparent f-value of the lens would change by crop factor.
APS-H would be f4 x 1.3 = f5.2
APS-C would be f6 or f6.4
Mi4/3 would be f8

Further the intensity of light transmission PER UNIT AREA would be constant, but because it would be falling on smaller targets, it would have a lower TOTAL light transmission - which is what would impact the total image brightness.

The quoting in your former post is not good. Two posts are mixed together. Not all mine.

You must realize the f-number is important for 2 things, maybe more but that I'm not aware off now. First it's a unit for exposure, secondly it is used as a simplified item for the dof. You must distinguish between them.

Let me try to formulate what I think you would like to know. One physical lens used on 3 different camera's with 3 different sensor sizes.
In that case you'll get 3 different images. For the exposure the F-number stays the same. The "light energie" is per m2 or something like that. A FF lens will only have a bigger image circle, a smaller sensor will only use a smaller part of it. The collected light per surface won't change, only the surface or sensor.

The other issue is the dof. When using that you must first define what is sharp. The accepted definition is you can tell the difference between a circle and a point. It's said that is circle of 0.3mm viewed at a distance of 1 meter is seen as a point. If the figures are different, sorry. Dof is based on a print of about 20x30cm and viewed at arm length. Knowing this the coc can be calculated for different sensor sizes. Viewing on your monitor will give another result as viewing from a print.

George

[Tronhard]

George you have obviously been responding while I was moaning about the total screw-up of my post. So first of all sorry that happened but I didn't write it that way. Hopefully my desperate efforts to correct this, despite delays and failures to post has finally got that one right.

The discussion AT THE MOMENT is about equivalence F-values compared to the values printed on a lens nor more no less. Is there such as thing as equivalence in f-values, where one multiples by the crop factor or not? THAT and ONLY that right now is what I am trying to sort out. Sure there are other questions but maybe one at a time please. As I said before I am looking for comments backed up by authoritative references on this subject.

[george013]

....
So, this is how it was presented in one submission on another forum:
"The aperture is specified as an F-number N, defined by the well-known formula

N = f/D

where f is the focal length of the thin lens (a.k.a. the effective focal length[SIZE=2] of the original lens assembly), and D is the diameter of the diaphragm opening." - quoted from lecture notes, Prof. LeVoy.

Note the word Effective, because if we place the lens on a crop sensor, it's effective (if you will apparent) focal length (fe) increases because its field of view narrows by the crop factor of the sensor, yet D remains a constant. So that changes the equation

I don't know where you are quoting from. It says "focal length of the original lens assembly".

The effective focal length is something else. The nominal focal length of a lens is the image distance when viewing at infinity. A crop sensor only catches a smaller part of that image as a ff sensor. When focusing on closer subjects, the image distance grows. The effective focal length is the image distance. important for makro photography where the image distance is 2x focal length at a magnification of 1.

George

[george013]

George you have obviously been responding while I was moaning about the total screw-up of my post. So first of all sorry that happened but I didn't write it that way. Hopefully my desperate efforts to correct this, despite delays and failures to post has finally got that one right.

The discussion AT THE MOMENT is about equivalence F-values compared to the values printed on a lens nor more no less. Is there such as thing as equivalence in f-values, where one multiples by the crop factor or not? THAT and ONLY that right now is what I am trying to sort out. Sure there are other questions but maybe one at a time please. As I said before I am looking for comments backed up by authoritative references on this subject.

The f-number as a tool for exposure doesn't change. It's a ratio between focal length and aperture diameter. It's a property of the lens, physical.
It will be different the f-number being a tool for dof.

George

[Tronhard]

So, this is how it was presented in one submission on another forum "..." They are not my words George I'm just quoting.

If one thing has come clear from this whole experience, it is that unless one goes to an authoritative source, confusion reigns because of disconnects in communication and definition. People have been bandying about terms with cheerful abandon, hence why I am looking for material with good academic or technical authenticity.

Meanwhile I am still being driven mad trying to post - "unable to connect" errors. Maybe it's something to do with the 100-odd mm we got dumped on us this weekend.

[Tronhard]

So, since the physical structure of a lens does not change when you switch it between bodies, and I mean it's lens diaphragm, you would say there is no such thing as lens f-value equivalence?

[george013]

So, since the physical structure of a lens does not change when you switch it between bodies, and I mean it's lens diaphragm, you would say there is no such thing as lens f-value equivalence?

The physical structure also includes the focal length. But again, this is for exposure. Not for the dof.

George

[pnodrog]

No to the idea that "brightness" is a function of the total light gathered by the sensor. As you know an image of a white cat on a large black rug maybe considered bright but the total light gathered low.

The equivalent F value concept in what I now think is a foolish and confusing effort to try and compare the same DOF of a lens when used on different sized sensors. It impacts both on exposure and total light falling on the sensor. However I rank exposure as a far more basic and significant than DOF. The equivalent F value is wrong and misleading regarding exposure comparisons and a lens's basic light gathering ability. Novice camera purchasers and even more so for older ex film camera users new to digital photography will be misled by the term. It is an invalid way to describe a lens and manufacturers are very unlikely to ever use it in specifications.

I can understand why a camera reviewer is looking for a better way to compare cameras but equivalence (particularly regarding f-stops) is not the way to go. Angles of view or magnification (compared to normal vision) as used for binoculars would be a better approach. After all when I first took up 35mm photography I had a table that gave me the angle of view for different focal length lenses. It was only after a few years of using different lenses I got a good feel for it. Then I went to medium format...

[pnodrog]

So, since the physical structure of a lens does not change when you switch it between bodies, and I mean it's lens diaphragm, you would say there is no such thing as lens f-value equivalence?

YES certainly as far as I am concerned.

An outboard motor's basic specifications don't change just because you have put it on a different boat.

[george013]

One more.

The intensity of light is a function with its radiation. Somebody else may explain that. But the energie hitting the surface is x per m2 and per time.
The image of a lens is a circle. The exposure in that circle is equal all over the surface, when it's right.
The circle is the image circle. The squares are the different sensor sizes. As you see the exposure on all the different sizes are equal. Only the sizes differ. If you take a ff picture, you can make a crop in pp to simulate a dx or mft picture. That's where the name crop camera is coming from.
A dx lens has a smaller circle. But you could use it on a mft camera.



Paul,
To what do you say yes?

George

[pnodrog]

George. I was saying yes to the statement there is no such thing as f-value equivalence. Certainly not as a lens specification. It may have some relevance when used as a tool for cross referencing a lens when used with different sized sensors but it can just be so easily be misunderstood that it is not worth the risk. It is of exceptionally limited value to those that do understand it and could be very misleading to people who don't understand it.

[JohnRostron]

Trev, I think the reason that there is little (if any) scientific information about the equivalence of cropped and non-cropped lenses in terms of light intensity is that, to a scientist, the answer is so blatantly obvious. It is what a scientist would regard as a given (or even an axiom). If you look at the diagram in George's recent post, within certain limits, the light intensity, measured in candelas or whatever, reflects the light energy received per unit area. This light intensity will not depend in any way on the area being measured. True, a FF image would receive more light that an APS-C, but the individual sensors are measuring the light intensity just at a specific tiny pixel position. So, from a photographer's point of view, F2=F2=F2 no matter what the sensor size.

This does not, of course, address the other aspect, that of depth of field. I think that I have enough knowledge to write a technical answer to this, but I wouild guess that others have addressed this.

John

[William W]

Aside, only to address Dave's comment:

Surely a far better DoF test than a tight portrait (real world though that may be), would be to use one of those auto-focus accuracy checking devices. That should allow us to read off the numerical scale/tape measure exactly how much is 'in focus' and more easily compare the two shots.

Yes, for the technically minded, but consider that "better" can variable when applied as a qualifier to learning: for many High School Students the shoot out is fun; visual; easily understood that each camera has the same Subject Distance and each Head is framed the the same . . . that there is a difference in DoF is obvious, it doesn't matter what degree of difference that it is for to show that there is a difference.

WW