[Class A]

FoV is directly related to focal length. They are inseparable.

That's wrong in the context of discussing different formats.

A 35mm FF lens does not change its focal length when mounted on an APS-C camera, yet the recorded FOV changes. In other words, with respect to what is relevant, i.e., the recorded image, FOV and focal length are not "inseparable".

Do you see your mistake?

Another way of looking at it is using a property of focal length: magnification or the projection of the subject into the sensor. It's NOT about the image circle.

When talking about different formats, the size of the image circle is paramount. If you use too small a sensor and hence cut out much of the image circle, you lost signal samples, i.e., you will get more noise.

It really does not matter what focal length is used to create a certain image circle. If we want to make apples to apples comparisons then images we compare must have the same FOV anyhow, so it does not matter what actual FOV it is or what actual focal length was used.

All we care about is to always capture the full image circle.

Total light alone is totally missing the point.

What do you mean by "total light alone"?

What is there more than total light?

On the contrary, looking at intensity alone would be missing the point. Because if you look at intensity (exposure) without regard for the size of the image circle, you are missing the point of how many photons were captured.

Your point of view -- that exposure is the be all and end all of photography -- would be correct if we always printed our images at a size proportional to the size of the sensor we captured them with. So Q images would make tiny prints and 645 images would make large prints. Only then exposure would accurately predict image quality because the lower amount of light used for a Q image could only be viewed in a tiny print in which image noise would not be observable to the extent it is on a large print.

Point is, decreasing focal length does brighten an image.

No, not if you keep the f-ratio and format constant.

Decreasing focal length does not brighten an image either, if you keep the aperture diameter constant and change the registration distance and format accordingly. The intensity gets higher, but the overall image brightness (intensity x area) does not become larger.

Point is, decreasing focal length does brighten an image. It lets MORE light per unit area.

More light per unit area, i.e, higher intensity only equates to a brighter image, if you keep the image size constant.

When talking about different formats, the image sizes (image circles) change. So more light per unit area does not mean a brighter image, if the image size reduces accordingly. A small, high-intensity image circle may look brighter than a large, low-intensity image circle but in terms of the whole image they have the same brightness. You can record either version and when you print them to the same output size, they'll look exactly the same.

Heck, it lets more light in, period.

And this is exactly where you are 100% wrong.

Higher intensity does not equate to "more light". It only does if you don't change the image size. However, in our discussion image size changes. You are one of those people who equate longer focal lengths with larger image circles, right? If so, you must agree that decreasing the focal length reduces the size of the image circle. If you do, you'll realise that the higher intensity is just caused by the same light now being condensed in a smaller area. In other words, there is not "more light". Do you understand this now?

[dtmateojr]

^^^ Another one who does not understand ratio.

One last time:

If x/y equals z then every sensel receives exactly z number of photons.

I'm outta here. This is hopeless.

[Class A]

ROFL! Somebody does not understand the term RATIO.
Is 1/2 less than 2/4? I'll give you a hint that any child would understand: the numerator is the incoming light and the denominator is the sensor area. LULZ!!!

You got two things correct:

1. 1/2 = 2/4.
2. intensity = incoming_light / sensor_area.

Where you seem to struggle is that a 100x100 sensor is 4x larger than a 50x50 sensor.

Say we have 2500 lumen of incoming light. So,
sensor_small_intensity = 2500 lumen / (50mmx50mm) = 1 lumen/mm^2
sensor_large_intensity = 2500 lumen / (100mmx100mm) = 1/4 lumen/mm^2

So the same illumination results in two stops less intensity on the larger sensor.

Hence my statement "In fact, the larger the format, the lower the exposure must be." is correct, if we want to capture the same amount of incoming light.

People like you want to keep the intensity (exposure) the same and would hence increase the amount of incoming light by a factor of four (i.e. two stops). That would have a significant impact on image noise, and if the increase in light is achieved by widening the aperture, it would also have a significant impact on DOF.

Do you understand now that exposure is not the deciding factor when it comes to image quality when changing formats?

[Class A]

If x/y equals z then every sensel receives exactly z number of photons.

Wot?

You probably wanted to equate "z" with the amount of light received per pixel (i.e. the intensity). For realistic scenes it is not correct to assume that "every sensel receives exactly z number of photons", but I'll let that one slide.

Now what you forget (for the umpteenth time) is that looking at a single pixel does not make sense when you are concerned about image quality. For the latter the amount of pixels also matters.

You have never responded to my argument of "more pixels = more samples -> better quality". Try to refute it. You will not succeed.

I'm outta here.

Promise?

---

Last edited by Class A; 07-18-2015 at 05:01 PM.

[ScooterMaxi Jim]

I'm sure you had good intentions and tried to help the OP in a more direct manner.

The trouble is that the deeper discussions that ensued in the thread are the result of people knowing what they are talking about trying to dispel the myth that you should convert focal length in order to understand which FF lens would take the same images like a certain lens on APS-C, and that DOF is affected somehow as well, but that an f/2.8 lens will remain a f/2.8 lens for the purposes of exposure.

While technically accurate, stating that the exposure aspect does not change is very misleading, because in order to take an image on a FF camera that looks the same like one taken on an APS-C camera, we have to use less exposure.

The important image parameters DOF and SNR (noise levels) are determined by the total amount of light used which is proportional to exposure x sensor-area.

That's why stating that "f/2.8 = f/2.8“ is entirely unhelpful. The correct equivalent f-ratio the OP needs to consider is f/4.

Please read the DPReview article.

P.S.: You were right about t-stops, but this discussion was not what the OP was after.

Let's start with your referenced Richard Butler article from which much misunderstanding has been interpolated. Like you, Butler has shared only a partial understanding of what's going on, and poor assumptions have followed accordingly. Other less-known articles have provided a deeper understanding of issues related to DoF and circle of confusion. Depth of field is perceptual, and it is influenced by the amount of image magnification. Every 50mm f/2.8 image is going to have the same actual DoF, just as the light intensity at t-stop 2.8 on a specified area of a sensor is always the same. The added magnification of the image from a smaller sensor changes the perception, or you can move closer to the image to make an equivalent adjustment - and this also alters your perception of the image. If this wasn't so, you couldn't trust that a D-FA 50mm at f/2.8 shot on a K-3 would have the same exposure and DoF characteristics as the DA 50mm shot at f/2.8 on a K-3. They do have the same basic characteristics. When the FA is shot on FF it still will have the same character in the cropped portion of the image area. However, magnification and/or your distance to image to compensate for the cropped sensor size difference results in the change of perception.

The OP wants to understand whether or not APSc essentially requires slower shutter speeds to achieve the equivalent f/2.8 speed he understands from 35mm film days. The answer to that question essentially is "no." He did not request an understanding of DoF and aperture equivalence - that was information well beyond his inquiry.

[wombat2go]

Hi Scooter,

You are quite free to use, in a diatribe,
punctuation, and line breaks.

And add paragraphs
so old guys like me can follow along.

Regards!

[Class A]

Let's start with your referenced Richard Butler article from which much misunderstanding has been interpolated.

I did not see how you referred to that article later on, but please note that neither Richard Butler nor I can be held responsible for the misunderstandings that people develop.

Like you, Butler has shared only a partial understanding of what's going on, and poor assumptions have followed accordingly.

Are you saying that we only partially shared our understanding or that our understanding lacks depth?

If it is the first: There would be quite a bit more to say on the matter, but neither did the OP ask for more depth nor did any debate so far warrant to go into any more depth.

If it is the second: Please share what makes you think that my understanding lacks depth. If I made an error somewhere, I'll concede it. If I just did not say as much as could have been said, please see above.

Note that I did explain how DOF depends on output size and viewing conditions (as it was necessary to make a point).

Also note, with respect to overloading the OP with information that the only answer to the OP that was strictly necessary was audiobomber's first reply. But if you then have people perpetuate old myths ("f/2.8 = f/2.8 = f/2.8") and argue against correct answers, some more posts and deeper arguments become necessary (if you care whether someone walks away with nonsense in their head or not).

The OP wants to understand whether or not APSc essentially requires slower shutter speeds to achieve the equivalent f/2.8 speed he understands from 35mm film days.

I cannot find one post by the OP where he enquires about shutter speed, but on that subject please note that the "1/focal-length" rule from 35mm film days to ensure sharp images with hand holding was designed to work for FF cameras.

It cannot be applied for APS-C cameras without adaptation. The APS-C rule is "shutter_speed = 1/(1.5 x focal-length)".

There are two ways to arrive at this adapted law:

1. One can realise that APS-C has an enlargment factor that is 1.5 x higher than that for FF. Hence any motion blur will show 1.5 x stronger. This necessitates a corresponding increase in shutter speed.
2. You can continue to use the 35 film days law unaltered, but then you must work with FF-equivalent parameters. For example, you'd have to pretend you are working with a 75/4 lens instead of a 50/2.8. This FF-equivalent focal-length of 75mm will give you the right shutter speed when using the original rule.

The answer to that question essentially is "no."

Incorrect.

If the OP wants the same image quality that he'd get from an f/2.8 lens on FF with the 17-50/2.8 then first, he could not without incurring deeper DOF, but second, he would have to use a slower shutter speed to collect the same amount of photons that the f/2.8 on FF collects.

You could only get away with using "f/2.8" on APS-C with the same shutter speed as on FF, if you always printed your APS-C images 1.5 x smaller than the FF images. However, people don't print / view according to the capture format and anyway, any fair comparison must work with the same output size.

---

Last edited by Class A; 07-19-2015 at 12:44 AM.

[wombat2go]

but second, he would have to use a slower shutter speed to collect the same amount of photons that the f/2.8 on FF collects.

Hi Class A,
Do you have : Holst & Lomheim 2nd Ed.?

Can you take us through, for example,
at R1 = 1200mm for Mamiya RB67 fl=90mm compared to Pentax aps-c fl=28mm and then Olympus M43 fl=18mm,
using H & L Table 2-3 Standard Photometric Units and article 2.6 + etc?

I saw a bit of lax usage of units in this thread. As engineer, I have to see absolutely correct units all through.
Thanks

Edit
6x7 58 by 67.7 mm =3926.6 mm^2
aps-c 15.6 by 23.6 mm = 368 mm^2
m43 13 by 17.3 mm = 224 mm^2

[Lowell Goudge]

Of course you can (see below).

However, that does not mean you should.
You are knocking down a straw man when you question the point in replicating images with different formats. There are a number of applications for the equivalence principle (e.g., making informed choices about equipment) that I don't want to discuss here but surely the point is not to shoot test targets, replicate shots, or be on a forum rather than "out and shooting".


You are correct that you have to change lenses (to account for the difference in focal length required) and that the same f-stop won't yield the same DOF. However, you simply stop down on the larger format to achieve the same DOF.

It can be mathematically proven that you can take the same photograph with different format cameras and it can also be empirically demonstrated. Have you looked at the real world demonstrations by DPReview?

I think you missed the point . I was commenting on the signal to noise ratio when comparing formats using identical optical relievers in each format

It is a gross assumption that the signal to noise ratio will be the same at different ISO if you stop down to get the same DOF, in an attempt to get the exact same image cross each of the two sensors, therefore you cannot stop down to make the comparison. Because you need to change the ISO. Perhaps you can Change the shutter to compensate, but again it is an assumption that the signal to noise ratio remains unchanged for an exposure of half the light and twice the exposure time. For me, while it may be nice to theorize you can do this, in practice you can't.


This does not mean that you can't make a comparison quickly for field of view for example using the crop factor, but to try to achieve 100% equivelence is impossible.

For me,,the only thing to know is

Image size = subject size x focal length / distance

[audiobomber]

This does not mean that you can't make a comparison quickly for field of view for example using the crop factor, but to try to achieve 100% equivelence is impossible.

I disagree, and posted specific examples to demonstrate. In what way are my examples non-equivalent?
https://www.pentaxforums.com/forums/10-pentax-slr-lens-discussion/299360-equi...ml#post3315680

Falconeye's conclusion at the end of his white paper on equivalence:

5. Conclusion

Sensor size plays no rôle as long as different cameras use equivalent parameters which implies a scaling of focal length, F-stop and ISO sensitivity. This holds true for "ideal" cameras which fully exploit the laws of nature.

In any comparison of different cameras, parameters should first be made equivalent in order to conclude non-trivial statements. I.e., scaling focal length, but not F-stop and/or ISO sensitivity must lead to trivial statements.

With equivalent scaling, all comparisons then amount to deviations from the ideal nature of a camera. This includes cost for a given performance or performance for a given cost or available options on the market. It was explained that the sweet spot where cost optimizes performance crawls to increasing sensor sizes, as performance requirements increase or technological progress advances, both effects being independent but amplifying each other.

Eventually, it was shown that cameras with a larger sensor form a superset of cameras with a smaller sensor if the pixel pitch is kept constant. As long as two cameras are in the intersection of both sets, they produce images which are indistinguishable.
http://www.falklumo.com/lumolabs/articles/equivalence/index.html

[Class A]

Do you have : Holst & Lomheim 2nd Ed.?

No, sorry I don't.

Can you take us through, for example,...

What do you want me to do?

I don't know what is in "article 2.6" and "Table 2-3". Is "R1" the distance between camera and subject?

The lenses you listed look like the "normal" lenses for their respective formats. The Olympus lens is slightly too short, but for the FF format "50mm" is considered "normal" although it should really be "43mm". Again, the definition of a "normal focal length" assumes a standard viewing distance between photo and beholder. If this distance isn't exactly the diagonal of the image then another than the standard "normal" focal length would provide undistorted perspective for the beholder.

[wombat2go]

What do you want me to do?

Since this thread got unreasonably technical, it could be beneficial to take it to a conclusion.
The way I request it, the optics guys who differ on this thread could take us through the analytical calculations for, say, an extended subject at 1200mm from N2.

We know that analytical calcs are not accurate below about f/4, so perhaps the experts could run analytical calculations for us, of a little bright green nose of 25mm dia of Maxwell's deamon, reflecting sunlight at about 530 um, at 30 degrees off the paraxial, for the above 3 sensors that I mention above.

Of course, we don't know the sensor characteristics, because they are not published, so take it up to there.
That lack did not seem to inhibit all the pontification and huffing and puffing in the earlier posts on this thread.

It seems like an easy job, for example in a spreadsheet or macro.

If the mainstream references that I mentioned above are not available, then use others and cite them.

Should be easy challenge, maybe 2 hours work, less time than the experts spent blowing hot air here , I suspect.

[audiobomber]

Should be easy challenge, maybe 2 hours work, less time than the experts spent blowing hot air here , I suspect.

Then why don't you do it?

[wombat2go]

Well Dan, I do such calculations all the time for my hobbies
For example here is a .pdf from a crude one on depth of field , not related to this thread, for use on my own cameras.
https://app.box.com/s/7xqm62t3d0ejm9zxcbob4622hh8ryz68

But my comments here are:
When we search say "Pentax Equivalent Focal length" we get such as this thread
I see that some newcomers queries on PF are just concerns about rhetoric (often untrue) from the forums.

When I look at the posts and photos contributed by those on threads such as this
it is clear that they are skilled either techno or as photographers.

But the posts are of a very low standard - argue, shoot from the hip, repeat rhetoric etc.
The articulation is very poor. Often the content of posts is based on simple but incorrect intuition on a complex subjects
that can't be visualized, such as Depth of Field, noise at low light etc.

So we have members who don't even read, and treat it as a joke, make silly comments etc
I suppose I am guilty of that too.

We have other readers as I see who accept what becomes pseudo science, jump on the bandwagon of it.

That was the reason for my post above to members who are obviously skilled:
::- slow down, cut the rhetoric, think about it, provide correct information backed up by references.

Would that not be preferable to this?

[BrianR]

Well Dan, I do such calculations all the time for my hobbies
For example here is a .pdf from a crude one on depth of field , not related to this thread, for use on my own cameras.
https://app.box.com/s/7xqm62t3d0ejm9zxcbob4622hh8ryz68

You can lead the way in upping the overall forum standards by making your spreadsheets available as you produce them, along with a brief explanation of what you're up to. They don't need to be polished presentations, but I would love to see your calculations for sheets like your example and am willing to work through it to figure out what's what.