The Recent F/stop Controversy

Started 1 month ago | Discussions thread
mamallama
mamallama Forum Pro • Posts: 55,629
Re: The Recent F/stop Controversy

Michael Fryd wrote:

Bill Ferris wrote:

Michael Fryd wrote:

Bill Ferris wrote:

Michael Fryd wrote:

...

Thank you for responding and adding to your comments.

If we know the total light gathered but don't know if a smartphone sensor, a medium format sensor or something in-between in size was the mode for using pixels to capture that light, how are we in a better-informed state than knowing an image was made at f/4, but not knowing the scene brightness or shutter speed? The same total volume of light may correspond to a blown out, uniformly white image at base ISO for a smartphone while appearing reasonably exposed at base ISO on medium format. Even on the same format sensor, a too high ISO or a too low ISO can significantly impact and limit the quality of the image we're able to make with a fixed total amount of light.

What you are actually asking about is how a photographer can determine when he has run out of headroom for increasing total light captured.

No, that's not what I asked. You've stated, "total light gathered tells us useful information about the final image without knowing the specifics of either the pixel or sensor size." My question is, knowing only the total light gathered, how do we know of the image is properly exposed, underexposed or overexposed? If you want to answer within the context of ability to recover details in the highlights and shadows, fine. But we both know that, limited only to knowing the total light gathered, we can't know if details will be lost due to over- or underexposure.

You're still thinking in terms of film. Traditionally "Proper exposure" means that our exposure hit the sweet spot on the film's S-shaped response curve. Miss that spot, and the negative is too dense or too light to get a good image.

Digital doesn't have the same sort of response curve. There's a wide range of exposures that produce a good quality raw file.

However, if we want to pretend that digital acts like film, we need to pretend that there is a "correct exposure". Therefore we have to pick one. Generally we say that we have "correct exposure" when the image lightness of the camera produced preview looks good.

Of course, what we really are saying is that the "correct exposure" is one where the exposure is a good match for the selected ISO.

I'm going to suggest that "correct exposure" is the wrong terminology. The correct terminology is "correct ISO". With digital there is no need to set the ISO and them make the exposure match. With digital you pick aperture and shutter, and then pick the ISO that corresponds to that exposure.

So the question of whether a digital image is "properly exposed" doesn't make sense unless we pretend that digital acts like film.

Total light gathered does give us a reasonable idea as to how noisy the image will be. This is similar to shutter speed which affects motion blur, and aperture which affects depth of field.

The answer is that digital cameras contain very sophisticated computers and can tell us.

All we are talking about is changing the scale from light per unit area, to light collected relative to what the camera can handle.

Cameras have limits on how wide or narrow the aperture can be. How fast/slow the shutter can be. And how much/little light can be recorded.

A hundred years of film photography has trained us that the light limits are more significant than the other two. This is no longer the case with digital.

What you're talking about, is the ambient brightness of the scene. Or, in the absence of ambient illumination, illumination added to the scene by the photographer. I submit that, if there is no light and no available option for adding light, aperture diameter and shutter speed are irrelevant. It is literally impossible to make a photograph without electromagnetic radiation (typically visible light) to which the medium is sensitive. Light, is and will always be the first and most important consideration.

Yes, if there is no light there can be no picture. Similarly, a camera may not be able to capture a useable image if there is too much light.

There are also situations where the camera does not have enough depth of field to get the entire subject in focus, or doesn't have a fast enough shutter to prevent motion blur.

These are situations which lie outside the shooting envelope of the camera. For instance a particular shot may lie inside the shooting envelope of a full frame camera with a maximum shutter speed of 1/8000, but outside the shooting envelope of a camera that maxes out at 1/2000.

Within this context, I would say it is more accurate to describe, f-stop, as a concept adopted as a tool of photography than as a concept linked more closely to film or any other mode of containing the light-sensitive medium. The term has no different meaning to a photographer using a chemical emulsion (e.g. glass plate, tin type, film) than to a photographer using pixels (e.g. CCD or CMOS) to capture light. It has no different influence on exposure.

The tradition of describing apertures as the ratio is intended to make it easy to hit a particular target exposure (light per unit area) on the film.

Obviously, it tells us the light per unit area when shooting digital, or other light sensitive mediums.

This and the fact that, regardless of the medium or mode, the same scene brightness, shutter speed and f-stop deliver the same exposure, make f-stop more informative to the photographer. The same f-stop can be achieved using a plethora of combinations of focal lengths and aperture diameters. The same aperture diameter can, within the context of the same scene brightness and shutter speed, deliver a wide range of exposures...a small number of which will be to our liking while the vast majority will be failures.

Again, you are assuming that there is an advantage to the photographer to being able to get the same exposure on two different cameras.

Suppose I have a full frame camera, and I am getting a perfect image shooting a 50mm lens at f/4, 1/60 and ISO 400.

The camera breaks, and I put out my 2X crop body to continue shooting. Why would I want to match the full frame exposure (light per unit area) on the 2X crop body?

If I was shooting film, matching the exposure is critical. If I don't I won't have a useable negative.

If I am shooting digital, matching the exposure will give me different results. At the same exposure, the 2X crop body will give me a noisier result. At the same f/stop, the 2X crop body will also give me deeper depth of field, and more diffraction blur.

If I am not shooting film, why would I want to keep the same exposure on my backup camera? Why not use the exposure that gives me the same results? (which will be an exposure that has the same total light, and same aperture diameter).

My point is that with digital this implementation detail is not that important in terms of what we will see in the final image. Therefore we are better off with a system that backs out unneeded implementation details.

When freed from the need to be a slave to traditional S-shaped film response curve, we have the opportunity to elevate the priority of depth of field and the field of view.

What is the direct relationship between the total amount of light captured and either depth of field or field of view? How about the degree to which motion within the scene is frozen.

Assuming the same subject, any combination of focal length, sensor size, shutter speed and f/stop that yield the same angle of view, depth of field and motion stopping captures the same total light. This is independent of sensor size.

Depth of field and angle of view are understood within the context of focal length, aperture diameter and distance to subject. Scene brightness and total light captured are irrelevant.

Actually the camera factors are angle of view and aperture diameter. You don't need to know the focal length and f/stop in order to compute depth of field.

However, if you have focal length and f/stop you can use the more complicated formula that compute angle of view and aperture diameter, and then the depth of field.

Total light enters into it because if you have the same angle of view, same shutter speed, and same aperture diameter, you will also have the same total light (assuming the same subject).

This relationship strongly suggests that the underlying important factors are aperture diameter, angle of view, subject illumination, and shutter speed.

Knowing f-stop and focal length, one can easily determine aperture diameter. One can manage both angle of view and depth of field by making changes to f-stop, focal length and distance to subject. It's not necessary to know the aperture diameter. There is no control or display on the camera or lens thay will tell you the aperture diameter being used. Knowing the aperture diameter is not necessary to the task of controlling depth of field. It's not even all that relevant top angle of view.

Actually most lens descriptions list the formula for the aperture diameter. When you say that the focal length is 100mm, and the aperture diameter is 1/4 the focal length (normally written as "f/4") you are just using a complicated way of saying a 25mm aperture diameter.

With respect to f-stop, a particular setting provides some indication that a shallow or deep depth of field for that focal length was used or that a relatively wide or narrow aperture diameter for that focal length was used.

Yes. The combination of f/stop and focal length provide an indication of depth of field. That's because they provide the two factors you need to know in order to determine aperture diameter.

It's easier to just work directly from aperture diameter.

Distance to subject is also a relevant factor. To be clear, I'm not arguing that aperture diameter is not a factor determining depth of field. However, putting that characteristic front & center is made very difficult for most photographers by the absence of any camera or lens setting or display that tells a photographer what the aperture diameter is. How do propose making that number readily available? More important, what are photographers not able to accomplish as a result of not having that measurement readily available?

Yes, distance is a relevant factor, however, for the sake of simplicity, I am only listing the factors we are discussing.

In terms of the camera not telling the photographer the aperture diameter, that's generally not true. My lens is labeled 50mm, and my camera might tell me the aperture diameter is set 1/8 of that value.

I submit, there is no creative goal that cannot be achieved by virtue of not knowing the aperture diameter. Isn't that, ultimately, the deciding factor in any argument for change? For real change to be embraced and effected by those who are impacted by the change, there must be done tangible, positive benefit to themselves, the community or organization they're part of. Replacing f-stop with aperture diameter or exposure with total light captured yields no tangible, practical benefits.

You need to know (or have enough information to compute) the aperture diameter in order to make creative decisions about depth of field and overall image noise.

You don't need to know focal length, f/stop or sensor size. Nor do you need enough information to compute those values.

My point, is that either number in isolation isn't fully instructive. However, as f-stop is one of only two camera settings determining exposure and as it is directly under the control of the photographer, f-stop is - along with shutter speed - a control a photographer benefits from mastering. Mastering an understanding of f-stop has positive impacts both technically and creatively.

Yes. modern cameras are labeled in f/stops, and actual mm. Yet we routinely suggest that people think in terms of effective f/stops, and effective focal lengths.

I only suggest folks think in effective or equivalent terms when comparing different formats. For the purposes of doing or talking about photography with a specific camera and lens, there is no need to bring up equivalence as a consideration.

Yes. If you only use one system, and you won't ever use another system, there is no need to know how to compare your system to others.

However, that doesn't mean there is an advantage to adding implementation details to the explanation.

I am suggesting that instead of "effective f/stop" we use the more straightforward "aperture diameter."

What are the practical impacts and tangible, real world benefits of your proposal? How does a photographer get that number? Where on the camera or lens do they look to see the aperture diameter displayed? What creative outcome that we currently are unable to accomplish, becomes doable by making this number readily available? Which creative outcome that is presently difficult to accomplish, becomes much easier to achieve? What are the negative consequences of displaying an aperture diameter as opposed to an f-stop?

Well, we started with the question should we get rid of the f/stop. I think there is a good case to be made the we should deemphasize that terminology and emphasize aperture diameter instead.

Digital cameras have a lots of computational power and sophisticated displays. I think camera firmware should be helpful enough to display the aperture diameter.

The creative advantage is that it reduces the amount of mental calculations one needs to do.

Assume you move closer to your subject and change the focal length to maintain the same framing. If the aperture diameter remains the same, you get the same depth of field. If the f/stop remains the same, the depth of field changes.

While knowing the total amount of light can, within the context of the collection area of the medium, inform an understanding of whether or not we've captured enough light to make (in one's opinion) a quality image, that number doesn't translate to any specific depth of field. It doesn't inform our understanding of the degree to which action has been frozen. It doesn't indicate any particular angle of view.

Total amount of light informs us on the subject of shot noise.

We are able top make decisions about acceptable shot noise, now, using scene brightness, shutter speed and f-stop to control exposure, and ISO as an indicator of acceptable exposure.Aperture diameter and angle of view inform us on the subject of depth of field and subject framing.

Not really. That's not enough information. You don't have enough information to determine total light captured, so you can't predict how noisy the image will be.

ISO 400, f/4, 1/60 might result in a noisy image if you are using a smart phone, and a clean image of you are using full frame.

Exposure alone (light per unit area) is not enough to determine shot noise. You need to know the sensor area so you can consider total light captured.

Photographers already control depth of field using focal length, f-stop and distance to subject.

Not really. you need to add sensor size in there or you can't determine depth of field. You need to know the sensor size you you can convert f/stop to aperture diameter and the figure out the angle of view.

Individually none of these tell us the whole story. Yet if we know these we can predict the results without knowing sensor size or focal length.

We're already able to predict the results without giving aperture diameter or total light captured any consideration.

Only because you are using formula that take total light and aperture diameter into consideration. If you don't have enough information to determine diameter and total light, then you don't have enough information.

Now if you know diameter and total light, you don't need enough information to computer focal length.

If anything, I might suggest we go another direction. Rather than focus on the total amount of light captured, let's consider the light available from the subject. Without using a device to measure and quantify the available light, we can still have an informed sense of what the possibilities are. Outside on a bright, sunny day (an abundance of light from the subject), we can have confidence in being able to use a fast shutter speed to capture enough light from a soccer player to make quality images. Outside on a clear, dark, moonless night (minimal available light from the subject), we can have confidence in needing a fast f-stop and a slow shutter speed to capture enough light for a quality image in a single exposure.

If you want to look at the source, we need to look at subject illumination, frame size, and the needed depth of field.

However none of this tell us whether we need a "fast" f/stop. A "fast" f/2.8 lens may be overkill to capture the shot on a full frame, and not nearly enough on a smart phone.

The problem is that on the full frame the f/2.8 lens probably has a larger aperture diameter than the smart phone's f/2.8 lens.

On the contrary and as stated, seeing and making a rough evaluation of scene brightness goes a long way toward informing the choices of f-stop and shutter speed made by a photographer. An experienced photographer will notice a change to the quality of light and adjust f-stop or shutter speed - almost instinctively - to preserve a desired exposure. This happens every day, thousands of times a day all over the world.

Again, this is where we differ.

Why is preserving proper exposure so much more important than preserving proper depth of field and proper shutter speed?

Knowing the f/stop is only useful if we know the sensor size, so we can determine the aperture diameter.

The profound physical and performance differences between chemical emulsions and pixels along with the relative performance characteristics of the modes (glass plates, tin types, film stock, sensor formats) used to precisely positioning these light-sensitive media at the image plane introduce a host of issues relevant to the photographic process. Equivalence, is but one. Frankly, there are too many to adequately address in a collection of books, let alone a single online forum discussion thread.

However, f-stop, is a term having a clear definition. Any lack of clarity on that point in this post, is my failure; not a failure of the long-established and accepted definition of the term. Its meaning and role in determining exposure are the same, regardless of the medium or mode used in the photographic process.

The meaning of f/stop is clear. What changes with digital is the priority of hitting a particular light per unit area.

I'm not sure I agree with this last comment.

Over time using a variety of digital cameras, it's common to develop a personal sense of one range of ISOs indicating an optimal exposure and another ISO range indicating a minimum acceptable exposure. Admittedly, I have no sense at all how many photons I'm capturing from a subject. But using ISO as an indicator of scene brightness, I can have a pretty good idea if the resulting image will be to my liking.

To be clear, I'm using ISO as an indicator of exposure. It's an indicator of the light available in the scene and delivered (as determined by f-stop and shutter speed) to the sensor. Arguably, the greatest strength of giving f-stop a priority over total amount of light in this analysis, is that f-stop has the same meaning, regardless of medium or format, while the same total amount of light can translate to very different outcomes in those areas.

So, while pixels may be able to deliver a quality image at a lower exposure than film, we still have an appreciation of optimum and minimum levels of exposure.

Absolutely, humans are smart and capable. We can learn to use overly complicated systems and get good results. This does not mean that the system being used is the best.

I have a manual transmission car. I could label the 5 forward gears with colors rather than numbers. With a little practice I could get a feel for each gear and learn to drive the car. That doesn't mean it wouldn't be easier to label with traditional numbers.

Let's give a bit more thought to your manual vs automatic transmission analogy. I learned to drive manual transmission cars. They're fun, no doubt about it. But for the vast majority of drivers, an automatic transmission vehicle is much easier to drive. That's a tangible, real benefit that changed an industry. People were able to drive anywhere they wanted before the automatic transmission was adopted as the standard. Some folks might even argue that four on the floor was more fun and, in some situations, more charitable. But the fact that the automatic transmission made the task of driving much easier overwhelmed any reservations or resistance.

So, what is the great benefit to photography and photographers that will come from ending the f-stop and exposure talk, and replacing them with aperture diameter and total light captured?

Benefits include:

  • The explanation of how things work is simpler. It leaves out implementation details that are not important to the result.
  • It moves mental models away from exposure being paramount, and onto a more equal footing for shutter speed and aperture.
  • it is independent of sensor size, which makes it easier to teach or discuss amongst people who may be using different sensor sizes.

Wow!! I just signed a $15,000 contract with a professional wedding photographer for my daughter and I hope he has all this stuff down pat. Otherwise I will be disappointed in the results. 

 mamallama's gear list:mamallama's gear list
Panasonic Lumix DMC-GX85 +1 more
Post (hide subjects) Posted by
tex
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow