[leekil]

That sums it up nicely... it has everything to do with marketing, and very little to do with providing useful information so the average joe can make an informed decision.

I think you're going to the wrong extreme on this. You can only go so far trying to come up with an equivalence to compare two cameras with vastly different sensor sizes. There is a range in the middle where you can take roughly equivalent photos with them, but once you start comparing the depth-of-field capabilities at either end of the range there isn't really any point, especially given that the capabilities are so different there. It's best to just give the difference in general terms about, say, the DOF -- "the compact camera has tons of DOF", and leave it at that. In your example, talking about a 24-680mm / F18-35, at that point the difference between f18 and f36 doesn't really even matter that much, since it is so wide with a small sensor.

The other thing about trying to do equivalence is that 95% of the camera-buying public cares about the equivalence *not one bit*. They usually are not trying to compare any of those things, because they are just dealing with one camera, ever. It's only the people who are quite a bit more highly technical that care about it, and even then, it becomes confusing (as this same repeated discussion on every camera forum ever illustrates). I'm not even sure that marketing cares about it much now.

TL; DR -- Yes, there is some benefit to discussing equivalence in particular cases, but not enough for a universal change in lens labelling for all parameters for all lenses.

[angerdan]

When trying to find a fast lens for low light I came upon little discussion with regarding how much crop sensors influence depth of field and bokeh.
Am I correct that to match the popular DA 35/2.4, when use on an APS-C camera, would be no better than a 50/3.5 on a full frame?

The photographic world loves to talk about the 'equivalent' focal length, but equivalent aperture never seems to get mentioned.
Consider the lens on the Sony RX10. It is usually referred to as being 24-600mm / F2.4-4 Why is it not referred to as 24-600mm / F6-8?

You have to differentiate between f-stop and t-stop. F-stop is physically, t-stop too.
So the meaning of equavilent aperture is ambivalent, depending on the view. Either the resulting depth of field in a certain combination or in the other case just for exposure (value).
pentaxforums.com/articles/photo-articles/table-of-equivalent-focal-lengths.html

...when looking at a lens, the smaller sensor requires more light.

I have no experience with the RX10, but only the HX90 with 1/2.3 sensor. Thus, although the lens on the HX90 is actually 4.1-123mm / F3.5-6.4, to get the same 'performance' on a full frame would need a 24-680mm / F18-35.
To keep things like the rule of thumb working an equivalent ISO number would need to be used.

Smaller sensors don't require more light.
Everything depends on the pixel pitch, so if the size of a pixel stays the same, amount of light and signal noise doesn't change too.

You mix things up, performance of f-stop will always be the same, independend from the sensor size.
Same with focal lengh, only thing is that with higher crop factor much of the visile image circle is outside of the image sensor area.
letmaik.github.io/pixelpitch/index.html
clarkvision.com/articles/does.pixel.size.matter/

Back in the day of film there really was only one film size for the majority - 35mm. Problem now is that there are 4 competing sensor sizes for the masses - 1", 4/3, APS-C and 35mm.

Now instead of being able to judge the 'performance' of a camera simply by looking at the focal length and max aperture, one has to consider two additional numbers (sensor size and max usable ISO).
An adjusted aperature simply makes it easier (like adjusted focal length) to compare apples to oranges.

There has been and still are several film formats (which are still available) at the mass marcet at one time.
Film format - Wikipedia

Performance of a camere depend on the sensor and it's pixel pitch as well as the internal noise reduction techniques.
With film cameras it has been the same, you always know which ISO from which manufacturer/brand /model gives which amount of grain.
So everything's mostly the same, maximal usable ISO and f-stop of the lens. Now even with optical image stabilisation for every lens there's an advantage at some shutter speeds.

[Nicolas06]

When speaking of apperture, most of the time, we do not use the absolute apperture size, but the ratio. The ratio is useful because it is related to the density of light and exposure but has nothing to do with noise or dof.

Noise is related to the total quantity of light that hit the sensor. The bigger the sensor with the same apperture (so with same light density) the more total light is received and the less noisy the photo will appear with similar sensor technology.

Dof is more related to the physical size of the apperture. A 35mm at f/2.4 has a physical apperture of 14.5mm, almost the same physical apperture as a 50mm f/3.5 that is 14.28mm. (35/2.4 = 14.58..., 50/3.5 = 14.28...)

If you shoot in the same conditions with a K1 and an RX10 (same framing), and use the same apperture rati and shutter speed you'll get the same exposure and isos. That's true.

Tthe physical apperture is going to be 3 time smaller in diameter because the focal lens is 3 time shorter. So blur is going to be 3 time smaller in size, 9 time less in surface (http://asklens.com/howmuchblur/#compare-1x-50mm-f2.8-and-3x-16mm-f2.8-on-a-0.9m-wide-subject)

As the sensor surface is 9 time less, the total amount of light received is also 9 time less (about 3 stops) and so there going to be much more noise. Interrestingly, the RX10 III for example get a measured high iso score of 472 and the Pentax K1 a score of 3280 with a factor of rougly 7. That's not exactly nine but not far from it. You can shoot at 3280 iso on the K1 and get the same noise the RX10 get at iso 472.

[Sandy Hancock]

the smaller sensor requires more light.

Where did you get that from? The light intensity is the same, so the exposure is the same, regardless of sensor size.

If I put a full frame lens on my Q7 and make a correct exposure at a certain aperture, shutter speed and ISO, then put it on my KP and repeat with the same parameters, then put it on my K-1 and repeat again, all three images will be correctly exposed (as long as the light hasn't changed).

Just stop worrying about equivalence

[Adam]

It's not the sensor size the affects the DOF. DOF is an optical effect produced by the lens. It makes no difference what sensor the lens is in front of. What is different with larger sensors is that in order to get things big enough in the frame, the focus distance has to he smaller (closer) .... It is this that results in smaller DOF at the equivalent apertures.

Great concise explanation!

[CarlJF]

As the sensor surface is 9 time less, the total amount of light received is also 9 time less (about 3 stops) and so there going to be much more noise. Interrestingly, the RX10 III for example get a measured high iso score of 472 and the Pentax K1 a score of 3280 with a factor of rougly 7. That's not exactly nine but not far from it. You can shoot at 3280 iso on the K1 and get the same noise the RX10 get at iso 472.

A difference easily explained by the different sensor technologies used in the K-1 and RX10.

All the discussions around the "total amount of light" are true in theory. But they're all based on the important caveat that all sensors should be equivalent, having the same sensisitivity, pixel pitch, and N/R characteristics. Outside lab testing, and even then, this caveat is rarely if ever met. Thus making this equivalence much less relevant in the real world than some would want you to think. Even more that today, with modern cameras, noise isn't a problem unless needing very high ISOs. For most people working under normal shooting condition noise isn't that much of an issue anymore.

[Steve Grosvenor]

Where did you get that from? The light intensity is the same, so the exposure is the same, regardless of sensor size.

If I put a full frame lens on my Q7 and make a correct exposure at a certain aperture, shutter speed and ISO, then put it on my KP and repeat with the same parameters, then put it on my K-1 and repeat again, all three images will be correctly exposed (as long as the light hasn't changed).

Just stop worrying about equivalence

Poor choice of words on my part.

Generally the size of each pixel is smaller on a smaller sensor so is able to gather less light.

A certain threshold of light is required in order for the sensor to work optimally (avoid noise).

With regards to noise, the IQ from a good 1/2.3 sensor at ISO 1600 is comparable to that from a good FF at ISO 25,600. Hence my comment that a smaller sensor needs more light to produce an image of comparable quality.

---------- Post added 10-30-17 at 09:41 AM ----------

Great concise explanation!

Completely agree... very well worded explanation. However does not change the fact that regardless of the science behind the optical effect that with a smaller sensor, not unlike the concept of virtual focal length, a concept of virtual aperture is useful.

Consider the scenario of someone moving from 35mm (FF) to an 4/3.

They love both the DOF and FOV they get with their current 50mm F2.0 with their current camera.

Many don't know that what the need to be looking for is a 25mm F1.0 lens. The fact that the the focal length is often stated as an equivalent, but not the aperture, implies to many that they should be able to achieve the same photo with the 4/3 50mm equivalent F2.0 lens as the lens they currently use.

,"m":0}]Depth of Field, Angle & Field of View, & Equivalent Lens Calculator - Points in Focus Photography

[Nicolas06]

A difference easily explained by the different sensor technologies used in the K-1 and RX10.

All the discussions around the "total amount of light" are true in theory. But they're all based on the important caveat that all sensors should be equivalent, having the same sensisitivity, pixel pitch, and N/R characteristics. Outside lab testing, and even then, this caveat is rarely if ever met. Thus making this equivalence much less relevant in the real world than some would want you to think. Even more that today, with modern cameras, noise isn't a problem unless needing very high ISOs. For most people working under normal shooting condition noise isn't that much of an issue anymore.

This is all relative and depend of what you shoot. If you shoot common subjects in daylight, the issue of a small sensor isn't really the noise but mostly lack of dof control, difficulties to reproduce fine texture and subtle gradients but noise isn't really a problem. And the limit between a small and big sensor is not only fuzy but change depending of who you ask.

If you shoot in low light or need quite high shutter speeds (weddings, event, actually anything indoor, sports, wildlife), you'll benefit greatly to get both fast apperture and great performance at high iso and you'll want a larger sensor. For all such cases, I'd say APSC is still well worth it and FF is advised if you have the money.

And again acceptance level may be different from person to person. Personally, I consider the high iso performance to be great, as soon as with the necessary post processing, the image look perfect technically. And the high iso performance to be acceptable if the photo truely look great but you could notice if your are careful that there some noise, a bit lack or sharpness maybe and few small issues.

So basically for me max "perfect" iso for K3 is 800 and accetable iso limit arround iso 3200. I would not consider the RX10 as perfect even at base iso. Not for the noise but for other limitations of the sensor. But that's each to its own.

[leekil]

Many don't know that what the need to be looking for is a 25mm F1.0 lens. The fact that the the focal length is often stated as an equivalent, but not the aperture, implies to many that they should be able to achieve the same photo with the 4/3 50mm equivalent F2.0 lens as the lens they currently use.

The aperture is already a "virtual aperture" -- as Nicolas06 said, the aperture, as expressed by f-stop, is a ration, not an absolute aperture size.

Maybe what you are looking for is more of a DOF measurement, given that if you are talking about the aperture, you could mean the light-gathering capability *or* the depth of field.

[Steve Grosvenor]

The aperture is already a "virtual aperture" -- as Nicolas06 said, the aperture, as expressed by f-stop, is a ration, not an absolute aperture size.

Maybe what you are looking for is more of a DOF measurement, given that if you are talking about the aperture, you could mean the light-gathering capability *or* the depth of field.

I see your point. The comments by various others have helped clarify the topic for me. Thanks!

@@@@

When the only common 'sensor' was 35mm film, light gathering and depth of field for a given focal length were linked.

With a 35mm camera, for a given focal length, DOF was primarily a function of f-stop and distance to subject. Various rule of thumb could be applied... and for the most part one had a good feel how a lens should perform based on its specifications.

Now one has to also consider crop factor if one uses more than one camera with different sensor sizes.

Using the Sony RX10 as an example, I knew what to expect from the camera with regards to field of view from the 'equivalent' focal length. But I was expecting better control over DOF considering the stated aperture of the lens.

I have perhaps babbled on too long on this pet peeve...

The 'stated' specifications are misleading with regard to what photos you are able to achieve with the camera.

Camera manufacturers (along with most reviewers) were quick to emphasize the benefits of smaller sensors, smaller less expensive lenses with similar reach, by providing 'equivalent focal length'.

Doing something similar with stated aperture would be similarly useful... but perhaps it is just a case of being hard for this old dog to learn new tricks.

As for this 'breaking' classic rule of thumb methods, do those regarding setting speed manually still make much sense?

Does it not make more sense to select the exposure speed based purely upon the degree one wishes to freeze action, aperture based on desired depth of field, then use sensor sensitivity (ISO) to adjust exposure, as most allow settings from 100 to 3200 (or higher) with little degeneration in image quality?

[UncleVanya]

Look - we had this in film too. 2.25x2.25, 6x7, 8x10, 4x5, 110 - all of these had different standard lens sizes and different depth of field due to apparent angle of view and distance to subject for a given focal length. BUT film was largely the same across all of these so there were no sensor related components to the discussion. Also few of these could swap lenses out across formats so lenses on a given platform were largely treated in abstraction - some rules of thumb about how focal lengths were similar were used (equivalence) but it was less controversial. Overall equivalence is useful as a tool for your mind to help grasp what works across platforms and help you make informed decisions about what lenses to use at the start of a new platform. Over time the native use supplants any need for equivalence and you figure out what works in what situation like you did on the first format. Don't fret and worry over accuracy of equivalence - it is a heuristic designed to give some decision making benefits to those working in multiple platforms and/or new to a platform.

[CarlJF]

Doing something similar with stated aperture would be similarly useful... but perhaps it is just a case of being hard for this old dog to learn new tricks.

There's no need to learn any new tricks, because there isn't any. You already mentioned that you know how to control DOF by chnaging the aperture. This is still the case with any camera. No matter the format, you already know that fully opening the aperture will give you the minimum DOF for this specific camera lens combo. There's no need to translate this into FF, or any format, equivalence. Because if the DOF is okay for what you want to achieve, you don't care about equivalence. And if it's not enough, calculating the equivalence will not help you to get a thinner DOF with this specific equipment...

It's better and more productive to know your equipment and what it can or can't do than calcuting equivalence relative to some gear you don't have or may not even exist...

[photoptimist]

I see your point. The comments by various others have helped clarify the topic for me. Thanks!

@@@@

When the only common 'sensor' was 35mm film, light gathering and depth of field for a given focal length were linked.

With a 35mm camera, for a given focal length, DOF was primarily a function of f-stop and distance to subject. Various rule of thumb could be applied... and for the most part one had a good feel how a lens should perform based on its specifications.

Now one has to also consider crop factor if one uses more than one camera with different sensor sizes.

Using the Sony RX10 as an example, I knew what to expect from the camera with regards to field of view from the 'equivalent' focal length. But I was expecting better control over DOF considering the stated aperture of the lens.

I have perhaps babbled on too long on this pet peeve...

The 'stated' specifications are misleading with regard to what photos you are able to achieve with the camera.

Camera manufacturers (along with most reviewers) were quick to emphasize the benefits of smaller sensors, smaller less expensive lenses with similar reach, by providing 'equivalent focal length'.

Doing something similar with stated aperture would be similarly useful... but perhaps it is just a case of being hard for this old dog to learn new tricks.

As for this 'breaking' classic rule of thumb methods, do those regarding setting speed manually still make much sense?

Does it not make more sense to select the exposure speed based purely upon the degree one wishes to freeze action, aperture based on desired depth of field, then use sensor sensitivity (ISO) to adjust exposure, as most allow settings from 100 to 3200 (or higher) with little degeneration in image quality?

It makes a lot of sense. One might even rename it "deptherture". But it only works for cameras without interchangeable lenses.

As soon as a lens can be moved from camera to camera (which may differ in format and resolution), the appearance the images taken with that one lens differ in both the DoF in the whole picture level (DoF in a fixed 8x10 print) and the DoF in the 100% pixel-peep level (DoF in a 300 ppi full-size print).

Moreover, if one crops an image in post or builds up a larger image (panoramas or Brenizer), then the equivalent deptherture would change. The point is that although DoF measured in microns is strictly a function of aperture and focal length, the effects of it on the photograph (measuring blur either as a percent of image size or in numbers of pixels) is a function of four numbers: focal length, aperture, final image size, and pixel size.

[UncleVanya]

It makes a lot of sense. One might even rename it "deptherture". But it only works for cameras without interchangeable lenses.

As soon as a lens can be moved from camera to camera (which may differ in format and resolution), the appearance the images taken with that one lens differ in both the DoF in the whole picture level (DoF in a fixed 8x10 print) and the DoF in the 100% pixel-peep level (DoF in a 300 ppi full-size print).

Moreover, if one crops an image in post or builds up a larger image (panoramas or Brenizer), then the equivalent deptherture would change. The point is that although DoF measured in microns is strictly a function of aperture and focal length, the effects of it on the photograph (measuring blur either as a percent of image size or in numbers of pixels) is a function of four numbers: focal length, aperture, final image size, and pixel size.

Forgive me but isn't distance to subject relevant? I know telephoto compression is a function of purely visual field (if I have sufficient resolution and crop a smaller chunk of a wider fov then blow it up - the compression appears in the blow up frame to be the same as the one natively taken with the telephoto - but I thought subject distance from the lens was a key part of DOF.

[leekil]

One might even rename it "deptherture".

Or one might not!

---------- Post added 10-31-17 at 10:51 AM ----------

Does it not make more sense to select the exposure speed based purely upon the degree one wishes to freeze action, aperture based on desired depth of field, then use sensor sensitivity (ISO) to adjust exposure, as most allow settings from 100 to 3200 (or higher) with little degeneration in image quality?

You can do that with TAv mode on your Pentax camera, though many people might prefer not to, as though the image quality may be quite good at 3200, they still prefer the increased dynamic range at 100. But it really all depends on which technical photographic factors you are most interested in, both for a particular photo, and in general for your photography.

---------- Post added 10-31-17 at 10:55 AM ----------

This is all relative and depend of what you shoot. If you shoot common subjects in daylight, the issue of a small sensor isn't really the noise but mostly lack of dof control, difficulties to reproduce fine texture and subtle gradients but noise isn't really a problem. And the limit between a small and big sensor is not only fuzy but change depending of who you ask.

It *is* all relative to what you shoot. And one of my points was that most people might not care about DOF control (or even know what it is).

If you shoot in low light or need quite high shutter speeds (weddings, event, actually anything indoor, sports, wildlife), you'll benefit greatly to get both fast apperture and great performance at high iso and you'll want a larger sensor. For all such cases, I'd say APSC is still well worth it and FF is advised if you have the money.

That's right, and those situations are also ones in which comparing "equivalent" apertures doesn't always make sense, since they won't really be equivalent.